Road conditions, often degraded by insufficient maintenance or adverse weather, significantly contribute to accidents, exacerbated by the limited human reaction time to sudden hazards like potholes. Early detection of distant potholes is crucial for timely corrective actions, such as reducing speed or avoiding obstacles, to mitigate vehicle damage and accidents. This paper introduces a novel approach that utilizes perspective transformation to enhance pothole detection at different distances, focusing particularly on distant potholes. Perspective transformation improves the visibility and clarity of potholes by virtually bringing them closer and enlarging their features, which is particularly beneficial given the fixed-size input requirement of object detection networks, typically smaller than the raw image resolutions captured by cameras. Our method automatically identifies the region of interest (ROI)—the road area—and calculates the corner points to generate a perspective transformation matrix. This matrix is applied to all images and corresponding bounding box labels, enhancing the representation of potholes in the dataset. This approach significantly boosts detection performance when used with YOLOv5-small, achieving a 45.7\% improvement in average precision (AP) at IoU thresholds of 0.5 to 0.95 for a single class, and notable improvements of 30.3\%, 78.6\%, and 278\% for near, medium, and far pothole classes, respectively, after categorizing them based on their distance. This work is the first to employ perspective transformation specifically for enhancing the detection of distant potholes.

Accurate measurement of pedaling kinetics and kinematics is vital for optimizing rehabilitation, exercise training, and understanding musculoskeletal biomechanics. Pedal reaction force, the main external force in cycling, is essential for musculoskeletal modeling and closely correlates with lower limb muscle activity and joint reaction forces. However, sensor instrumentation like 3-axis pedal force sensors is costly and requires extensive post-processing. Recent advancements in machine learning (ML), particularly neural network (NN) models, provide promising solutions for kinetic analyses. In this study, an NN model was developed to predict radial and mediolateral forces, providing a low-cost solution to study pedaling biomechanics with stationary cycling ergometers. Fifteen healthy individuals performed a 2-minute pedaling task at two different self-selected (58±5 rpm) and higher (72±7 rpm) cadences. Pedal forces were recorded using a 3-axis force system. The dataset included pedal force, crank angle, cadence, power, and participants' weight and height. The NN model achieved an inter-subject normalized root mean square error (nRMSE) of 0.15±0.02 and 0.26±0.05 for radial and mediolateral forces at high cadence, respectively, and 0.20±0.04 and 0.22±0.04 at self-selected cadence. The NN model's low computational time suits real-time pedal force predictions, matching the accuracy of previous ML algorithms for estimating ground reaction forces in gait.

As the use of AI in medical imaging has increased, so has the need to explain a model’s results. Segmentation models are one technique used to produce explainable results. Due to larger size and sophistication, segmentation models which operate on 2D data can often produce better results than models operating on 3D data. In the real world, imaging is combined with clinical factors for diagnosis. To replicate this, multimodal models are used which combine image and text modalities. I propose a multimodal arrangement that converts 3D scans to 2D and uses a 2D segmentation model (DeepLabV3) to analyze the images. This is combined with clinical biomarkers to achieve a complete confidence score. Using the Medical Segmentation Decathlon Lung (MSDL) dataset and the LUng CAncer Screening dataset (LUCAS), I achieved a testing Dice coefficient of 0.91 on segmentation with a receiver operating characteristic (ROC), average precision (AP), and F-Score of 0.89, 0.91, and 0.85 respectively on the final multimodal results. This approach raises the F-score from the previously achieved value of 0.508 to 0.85, creating a new baseline of what is achievable with multimodal lung cancer diagnostics and possible methods to achieve these accuracies.

The objective of person re-identification (ReID) tasks is to match a specific individual across different times, locations, or camera viewpoints. The prevalent issue of occlusion in real-world scenarios affects image information, rendering the affected features unreliable. The difficulty and core challenge lie in how to effectively discern and extract visual features from human images under various complex conditions, including cluttered backgrounds, diverse postures, and the presence of occlusions. Some works have employed pose estimation or human keypoint detection to construct graph-structured information to counteract the effects of occlusions. However, this approach introduces new noise due to issues such as the invisibility of keypoints. Our proposed module, in contrast, does not require the use of additional feature extractors. Our module employs multi-scale graph attention for the reweighting of feature importance. This allows features to concentrate on areas genuinely pertinent to the re-identification task, thereby enhancing the model's robustness against occlusions. To address these problems, a model that employs multi-scale graph attention to re-weight the importance of features is proposed in this study, significantly enhancing the model’s robustness against occlusions. Our experimental results demonstrate that, compared to baseline models, the method proposed herein achieves a notable improvement in mAP on occluded data sets, with increases of 0.5%, 31.5%, and 12.3% in mAP scores.

Machine Learning (ML) has revolutionized healthcare by enhancing diagnostic capabilities because of its ability to analyze large datasets and detect minor patterns often overlooked by humans. This is beneficial, especially in pain recognition, where patient communication may be limited. However, ML models often face challenges such as memorization and sensitivity to adversarial examples. Regularization techniques like mixup, which trains models on convex combinations of data pairs, address these issues by enhancing model generalization. While mixup has proven effective in image, speech, and text datasets, its application to time-series signals like electrodermal activity (EDA) is less explored. This research uses ML for pain recognition with EDA signals from the BioVid Heat Pain Database to distinguish pain by applying mixup regularization to manually extracted EDA features and using a Support Vector Machine (SVM) for classification. Results showed that this approach achieved an average accuracy of 75.87% using leave-one-subject-out cross-validation (LOSOCV) compared to 74.61% without mixup. This demonstrates the mixup’s efficacy in improving ML model accuracy for pain recognition from EDA signals. This study highlights the potential of mixup in ML as a promising approach to enhance pain assessment in healthcare.

Diffusion Models (DMs) are probabilistic models that create realistic samples by simulating the diffusion process, gradually adding and removing noise from data. These models have gained popularity in domains such as image processing, speech synthesis, and natural language processing due to their ability to produce high-quality samples. As DMs are being adopted in various domains, existing literature reviews that often focus on specific areas like computer vision or medical imaging may not serve a broader audience across multiple fields. Therefore, this review presents a comprehensive overview of DMs, covering their theoretical foundations and algorithmic innovations. We highlight their applications in diverse areas such as media quality, authenticity, synthesis, image transformation, healthcare, and more. By consolidating current knowledge and identifying emerging trends, this review aims to facilitate a deeper understanding and broader adoption of DMs and provide guidelines for future researchers and practitioners across diverse disciplines.

Based on the detailed exploration of AI techniques for the early identification of learning difficulties in education, this study underscores the pivotal role of educational data analysis in enhancing student outcomes. Educators can proactively identify and address learning challenges through systematic data collection and analysis, thereby minimizing long-term academic obstacles and emotional stress. The integration of AI, particularly through natural language processing and computer vision technologies, facilitates personalized learning paths and real-time interventions, offering tailored support that maximizes each student's learning potential. These advancements optimize educational resource allocation and foster a more inclusive and effective learning environment, paving the way for future advancements in personalized education and intelligent teaching practices.

The rapidly evolving domain of multimodal sentiment analysis is crucial for unraveling the intricate layers of emotional expression in social media content, customer service interactions, and personal vlogs. This research introduces a cutting-edge Advanced Cross-Modal Gating (ACMG) framework that significantly enhances the precision of sentiment analysis by refining the interplay among textual, auditory, and visual modalities. Our approach addresses three foundational aspects of sentiment analysis: (1) Advanced learning of cross-modal interactions, which focuses on extracting and synthesizing sentiment from varied modal inputs, thus providing a holistic view of expressed emotions; (2) Mastery over the temporal dynamics of multimodal data, enabling the model to maintain context and sentiment continuity over extended interactions; and (3) Deployment of a novel fusion strategy that not only integrates unimodal and cross-modal cues but also dynamically adjusts the influence of each modality based on its contextual relevance to the sentiment being expressed. The exploration of these dimensions reveals that the nuanced modeling of cross-modal interactions is crucial for enhancing model responsiveness and accuracy. By applying the ACMG model to two highly regarded datasets—CMU Multimodal Opinion Level Sentiment Intensity (CMU-MOSI) and CMU Multimodal Opinion Sentiment and Emotion Intensity (CMU-MOSEI)—we achieve groundbreaking accuracies of 83.9\% and 81.1\%, respectively. These results represent significant improvements of 1.6\% and 1.34\% over the current state-of-the-art, showcasing the superior performance and potential of our approach in navigating the complexities of multimodal sentiment analysis.

Many systems in the manufacturing industry have spatial distribution characteristics, which correlate with both time and space. Such systems are known as distributed parameter systems (DPS). Due to the spatiotemporal coupling characteristics, the modeling of such systems is quite complex. The paper presents a new approach for three-dimensional fuzzy modeling using Simultaneous Perturbation Stochastic Approximation (SPSA) for nonlinear DPS. The Affinity Propagation clustering approach is utilized to dynamically determine the optimal number of fuzzy rules and construct a collection of preceding components for three-dimensional fuzzy models. Fourier space base functions are used in the resulting components of three-dimensional fuzzy models, and their parameters are learned by the SPSA algorithm. The proposed three-dimensional fuzzy modeling technique was utilized on a conventional DPS within the semiconductor manufacturing industry, with the simulation experiments confirming its efficacy.

Neural network regularized learning has garnered significant attention in recent years. We give a systematic investigation on the performance of regularized regression associated zonal translation networks. We propose the concept of Marcinkiewicz-Zygmund inequality Setting (MZIS) for the scattered nodes collected from the unit sphere. We show that, under the MZIS, the corresponding convolutional zonal translation network has reproducing property. Based on these facts,we propose a kind of kernel regularized regression learning framework and provide upper bound estimate for the learning rate with the kernel approach. We also give proof for the density of the zonal translation network with spherical Fourier analysis.We provide the approximation error with a K-functional.

Recent studies have shown correlations between the microbiota's composition and various health conditions. Machine learning (ML) techniques are essential for analyzing complex biological data, particularly in microbiome research. ML methods help analyze large datasets to uncover microbiota patterns and understand how these patterns affect human health. This study introduces a novel approach combining statistical physics with the Monte Carlo (MC) methods to characterize bacterial species in the human microbiota. We assess the significance of bacterial species in different age groups by using notions of statistical distances to evaluate species prevalence and abundance across age groups and employing MC simulations based on statistical mechanics principles. Our findings show that the microbiota composition experiences a significant transition from early childhood to adulthood. Species such as Bifidobacterium breve and Veillonella parvula decrease with age, while others like Agathobaculum butyriciproducens and Eubacterium rectale increase. Additionally, low-prevalence species may hold significant importance in characterizing age groups. Finally, we propose an overall species ranking by integrating the methods proposed here in a multicriteria classification strategy. Our research provides a comprehensive tool for microbiota analysis using statistical notions, ML techniques, and MC simulations.

In manufacturing chatter is an unwanted phenomenon that can lead to product quality reduction and tool wear. Real time chatter detection is key to preventing these issues and improving overall machining efficiency. In this paper we propose an unsupervised chatter detection method using autoencoders (AE) and Density-Based Spatial Clustering of Applications with Noise (DBSCAN) clustering algorithm that uses internal signals of Computer Numerical Control (CNC) machines. The proposed method starts by using an AE to extract features from raw internal signals collected from CNC machines. This step reduces the dimensionality of the data and captures the underlying patterns of chatter. Then the extracted features are fed into DBSCAN clustering algorithm which is a density based algorithm that groups similar data points and identifies outliers. We tested the proposed method with real world data collected from various CNC machines. The results show that our unsupervised chatter detection method has high accuracy, precision and recall, can detect chatter and distinguish it from normal machining. Also the method is robust to noise and can adapt to dynamic machining conditions. In summary our work presents an unsupervised chatter detection method using AE and DBSCAN clustering that uses internal signals of CNC machines. This method is a reliable and efficient solution for real time chatter detection so manufacturers can improve product quality, optimize machining process and reduce tool wear during machining.

Objective: Interstitial fibrosis / tubular atrophy (IFTA) is a common, irreversible, and progressive form of chronic kidney allograft injury, and it is considered a critical predictor of kidney allograft outcomes. The extent of IFTA is estimated through a graft biopsy, while a non-invasive test is lacking. The aim of this study was to evaluate the feasibility and accuracy of an MRI radiomic-based machine learning (ML) algorithm in estimating the degree of IFTA in a cohort of transplanted patients. Approach: Patients who underwent MRI and renal biopsy within a 6-month interval from 1/1/2012 to 1/3/2021 were included. Stable MRI sequences were selected, and renal parenchyma, renal cortex and medulla were segmented. After image filtering and pre-processing, we computed radiomic features which were subsequently selected through a LASSO algorithm for their highest correlation with the outcome and lowest intercorrelation. Selected features and relevant patients’ clinical data were used to produce ML-algorithms using 70% of the study cases for feature selection, model training and validation with a 10-fold cross-validation, and 30% for model testing. Performances were evaluated using AUC with 95% confidence interval. Main results: 70 coupled tests (63 patients, 35.4% females, mean age 52.2 years) have been included and subdivided into a wider cohort of 50 for training and a smaller cohort of 20 for testing. For IFTA ≥ 25%, AUCs in test cohort were 0.60, 0.59, and 0.54 for radiomic features only, clinical variables only, and combined radiomic-clinical model, respectively. For IFTA ≥ 50%, AUCs in training cohort were 0.89, 0.84, 0.96, and in test cohort were 0.82, 0.83, and 0.86, for radiomic features only, clinical variables only, and combined radiomic-clinical model, respectively. Significance: An ML-based MRI radiomic algorithm showed promising discrimination capacity for IFTA>50%, especially when combined with clinical variables. These results need to be confirmed in larger cohorts.

The detection and recognition of tables in image-based documents is a challenging and essential task for automated data extraction and anal- ysis. This paper introduces a methodology that utilizes Convolutional Neural Networks (CNNs) to address this problem. By harnessing the ro- bust visual recognition abilities of CNNs, the proposed method effectively detects and extracts tables from image-based documents. Additionally, an Encoder-Dual Decoder (EDD) architecture is employed to extract the textual content from table cells, enabling the transformation of visual data into structured, machine-readable formats. This structured data is then further processed using a Vector Space Model (VSM)-based language modeling technique for query-based table retrieval. The methodology’s accuracy is assessed by analyzing the spatial relationships between the extracted cells and comparing them against a pre-defined table structure. The results provide an evidence to the effectiveness of the proposed ap- proach, with an impressive 85% accuracy rate and a corresponding error rate of 15%. This accurate extraction and recognition of tables from image-based documents can significantly enhance applications in data analysis, information retrieval, and knowledge extraction.

Video self-supervised learning (VideoSSL) offers significant potential for reducing annotation costs and enhancing a wide range of downstream tasks in video understanding. The ultimate goal of VideoSSL is to achieve human-level video intelligence across a spectrum of tasks, from low-level tasks such as pixel temporal correspondence to high-level complex spatio-temporal tasks like action recognition. However, most existing VideoSSL methods focus on isolated aspects of this spectrum and fail to integrate different levels of task complexity. Our study presents the first comprehensive survey that connects all families of VideoSSL methods. We provide a detailed review of the full spectrum of VideoSSL, from low to high levels, by conceptually linking their self-supervised learning objectives and including a comprehensive categorization. Our extensive evaluation highlights the strengths and limitations of each SSL objective across various downstream task families. We also detail the challenges in current VideoSSL research such as data curation, interpretability, deployment, and privacy concerns, an area that previous surveys have not thoroughly explored. In addressing these challenges, we recognize the strengths of existing methods in addressing these challenges and outline future directions for research.

The accelerated development of AI technology has brought about revolutionary changes in various fields of society. Recently, it has been emphasized that Fairness, Accountability, Transparency, Ethics (Explainability) (FATE) should be considered to support the reliability and validity of AI-based decision-making. However, in the case of autonomous driving technology, which is directly related to human life and requires real-time adaptation and response to various changes and risks in the real world, environmental adaptability must be considered in a more comprehensive and converged manner. In order to derive definitive evidence for each object in a convergent autonomous driving environment, it is necessary to transparently collect and provide various types of road environment information for driving objects and driving assistance, and to construct driving technology that is adaptable to various situations by considering all uncertainties in the real-time changing driving environment.This allows unbiased and fair results based on flexible contextual understanding, even in situations that do not conform to rules and patterns, by considering the convergent interactions and dynamic situations of various objects that are possible in a real-time road environment. The transparent, environmentally adaptive, and fairness-based outcomes provide the basis for the decision-making process and support clear interpretation and explainability of decisions. All of these processes enable autonomous vehicles to draw reliable conclusions and take responsibility for their decisions in autonomous driving situations. Therefore, this paper proposes an Adaptability, Fairness, Transparency, Explainability and Accountability (AFTEA) Framework to build a stable and reliable autonomous driving environment in dynamic situations. This paper explains the definition, role, and necessity of AFTEA in artificial intelligence technology and highlights its value when applied and integrated into autonomous driving technology. The AFTEA framework with environmental adaptability will support the establishment of a sustainable autonomous driving environment in dynamic environments, and aims to provide a direction for establishing a stable and reliable AI system that adapts to various real-world scenarios.

In recent years, the delivery of medical supplies has faced significant challenges due to natural disasters and recurrent public health emergencies. Addressing the need for improved logistics operations during such crises, this article presents an innovative approach integrating vehicle and Unmanned Aerial Vehicle (UAV) logistics to enhance the efficiency and resilience of medical supply chains. Our study introduces a dual-mode distribution framework that employs the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm for efficiently clustering demand zones unreachable by conventional vehicles, thereby identifying areas requiring UAV delivery. Furthermore, we categorize the demand for medical supplies into two distinct sets based on vehicle accessibility, optimizing distribution routes via both UAVs and vehicles. Through comparative analysis, our findings reveal that the artificial bee colony (ABC) algorithm significantly outperforms the genetic algorithm in terms of solving efficiency, iteration counts, and delivery speed. However, the ABC algorithm's tendency towards early local optimization and rapid convergence leads to potential stagnation in local optima. To mitigate this issue, we incorporate a simulated annealing technique into the ABC framework, culminating in a refined optimization approach that successfully overcomes the limitations of premature local optima convergence. The experimental results validate the efficacy of our enhanced algorithm, demonstrating reduced iteration counts, shorter computation times, and substantially improved solution quality over traditional logistic models. The proposed method holds promise for significantly improving the operational efficiency and service quality of the healthcare system's logistics during critical situations.

Frequently Asked Questions (FAQs) systems rely on semantic symmetry similarity measuring between two sentences. To address the challenges of insufficient training data and limited domain-specific understanding in low-resource FAQs, this paper proposes a general framework, PDAMF-FAQ, to solve these issues. Firstly, we propose a paraphrasing-based data augmentation model that integrates syntactic information and edit vectors. Using a rule-based approach, it retrieves template sentences from a corpus and masks relevant words with special characters. Edit vectors between the original and paraphrase sentences are added to a pre-trained model's encoding layer to enhance the model's ability to learn the differences between the original and reference paraphrase sentences. Additionally, this paper presents a mixed-feature semantic matching model based on SimBERT. The model extracts keyword features from the text, replacing these keywords with special characters to construct intent features. These intent features, along with the user question and keyword features, are then concatenated to form the model's input. Experiments were conducted respectively on the paraphrasing-based data augmentation model, mixed-feature semantic matching model and their comprehensive application in a low-resource domain-specific FAQ. The experimental results show that the proposed framework effectively improve the performance of domain-specific FAQ system.

The popularity of handheld devices for point-of-care ultrasound (POCUS) has increased in recent years due to their portability and cost-effectiveness. However, POCUS has the drawback of lower imaging quality compared to conventional ultrasound, because of hardware limitations. Improving the quality of POCUS through post-image processing would therefore be beneficial, with deep learning approaches showing promise in this regard. This review investigates the state-of-the-art progress of image enhancement using deep learning suitable for POCUS applications. A systematic search was conducted from January 2024 to February 2024 on PubMed and Scopus. From the 457 articles that were found, the full text was retrieved for 69 articles. From this selection, 15 articles were identified addressing multiple quality enhancement aspects. A disparity in the baseline performance of the low-quality input images was seen across these studies, ranging between 8.65–29.24 dB for the Peak Signal-to-Noise Ratio (PSNR) and 0.03-0.71 for the Structural Similarity Index Measure (SSIM). In six studies, where both PSNR and the SSIM metrics were reported for the baseline and the generated images a mean difference of 6.60 (SD ± 2.99) and 0.28 (SD ± 0.15) was observed for the PSNR and SSIM, respectively. The reported performances demonstrate the potential of deep-learning-based image enhancement for POCUS. However, variability in the extent of performance gain across datasets and articles was notable, and the heterogeneity across articles makes quantifying the exact improvements challenging.

This work introduces the mathematical framework of the novel “First-Order Comprehensive Adjoint Sensitivity Analysis Methodology for Neural Ordinary Differential Equations (1st-CASAM-NODE)” which yields exact expressions for the first-order sensitivities of NODE decoder-responses to the NODE parameters, including encoder initial conditions, while enabling the most efficient computation of these sensitivities. The application of the 1st-CASAM-NODE is illustrated by using the Nordheim-Fuchs reactor dynamics/safety phenomenological model, which is representative of physical systems that would be modeled by NODE while admitting exact analytical solutions for all quantities of interest (hidden states, decoder outputs, sensitivities with respect to all parameters and initial conditions, etc.). This work also lays the foundation for the ongoing work on conceiving the “Second-Order Comprehensive Adjoint Sensitivity Analysis Methodology for Neural Ordinary Differential Equations (2nd-CASAM-NODE)” which aims at yielding exact expressions for the second-order sensitivities of NODE decoder-responses to the NODE parameters and initial conditions while enabling the most efficient computation of these sensitivities.

Neural Fractional Differential Equations (Neural FDE) represent a neural network architecture specifically designed to fit the solution of a fractional differential equation to given data. This architecture combines an analytical component, represented by a fractional derivative, with a neural network component, forming an initial value problem. During the learning process, both the order of the derivative and the parameters of the neural network must be optimised. In this work, we investigate the non-uniqueness of the optimal order of the derivative and its interaction with the neural network component. Based on our findings, we perform a numerical analysis to examine how different initialisations and values of the order of the derivative (in the optimisation process) impact its final optimal value. Results show that the neural network on the right-hand side of the Neural FDE struggles to adjust its parameters to fit the FDE to the data dynamics for any given order of the fractional derivative. Consequently, Neural FDEs do not require a unique α value, instead, they can use a wide range of α values to fit data. This flexibility is beneficial when fitting to given data is required, and the underlying physics is not known.

Our world is increasingly challenged by managing the impacts of natural disasters, particularly floods, which are frequent, dangerous, and costly. Traditional flood mapping methods, reliant on Optical satellites (MSI), struggle under cloud cover which is typical during such events. Synthetic Aperture Radar (SAR) offers a promising alternative with its cloud-penetrating capability, though its use has been limited due to complexity and data labeling challenges. This project aims to develop a SAR-based flood segmentation model that can rapidly and accurately map floods globally, with high adaptability to different flood types and regions. By utilizing deep learning and a novel transfer-learning technique that combines the strengths of Optical/MSI and SAR data, the model seeks to bypass the challenges of manual labeling and improve mapping accuracy. Initial results show the model's effective generalization across various flood events, with superior performance indicated by an Intersection over Union (IoU) of 0.72, outperforming existing methods and demonstrating promising capabilities in precise flood mapping.

Artificial intelligence harbors significant, unrealized promise for educational reformers, adding benefit to low-resource settings, particularly in the Least Developed Nations where access to quality learning remains most acute. The purpose of the introduction is to summarize how AI can reduce the educational gap and increase the effectiveness of education in poor resource environments. This paper, in a critical way, reviews the existing literature on the integration of AI in the education systems of developing countries concerning the benefits and challenges. This is achieved by reviewing a cross-section of literature from MDPI, Taylor & Francis, Springer, Wiley Online Library, Science Direct, and IEEE Xplore databases. Much emphasis within the selection criteria is given to those studies that had either empirically or theoretically demonstrated the influence of AI on educational practices within developing countries. Key findings: AI has the potential to greatly improve learning outcomes; it somewhat increases access to education but bodes well for the personalization agenda. These, however, will be surmounted upon adoption if it is to be effective: there exist huge challenges such as limited access to the internet, infrastructural deficit, and socio-cultural barriers. This study can, therefore, give inference to the fact that AI can change education and further enhance new equity in developing countries if there are corresponding proper policy interventions, building the capacity of teachers, and involvement of the community. More research is required to inform the long-term implications of AI and new uses being developed for AI applications in many diverse educational contexts, on which to base strategies for more equitable and more inclusive global education.

This research explores the application of ARIMA (Autoregressive Integrated Moving Average) models and parallel computing techniques to analyze and forecast anemia diagnoses in children under 36 months in the Junín region of Peru. Using health data from 2023-2024, including insurance type, patient information, diagnosis dates, hemoglobin levels, and treatment details, we develop predictive models to understand trends and patterns in childhood anemia. The study aims to demonstrate the effectiveness of time series analysis and high-performance computing in addressing this critical public health issue. Results indicate improved forecasting accuracy and computational efficiency, potentially aiding in resource allocation and policy development for anemia prevention and treatment programs.

Recommender Systems (RecSys) are crucial in managing information overload and enhancing user satisfaction across various digital platforms, including e-commerce and entertainment. Evolving from traditional models to Deep Neural Networks (DNNs) and, more recently, Large Language Models (LLMs), these systems leverage sophisticated algorithms to analyze user behaviors and preferences. LLMs, such as GPT-4, are trained on extensive datasets to comprehend and generate natural language, significantly advancing their ability to deliver personalized recommendations. This tutorial explores the transformative impact of LLMs on RecSys, discussing their development, application in handling complex datasets, and the integration of contextual insights. Real-world examples illustrate how LLMs enhance recommendation accuracy and user experience, highlighting challenges and future directions in the field.

Glaucoma, a global leading cause of blindness, necessitates accurate segmentation of the optic disc (OD) and optic cup (OC) for effective screening. However, existing segmentation methods based on convolutional neural networks (CNNs) suffer from high computational complexity and long inference times. In this paper, we propose an end-to-end segmentation method for the OD and OC using a lightweight MobileNetv3 as the feature extraction network. Our approach incorporates boundary branches and an adversarial learning network to achieve multi-label segmentation of the OD and OC. We evaluate the proposed method on three publicly available datasets: Drishti-GS, RIM-ONE-r3, and REFUGE. The experimental results demonstrate segmentation accuracies of 0.974/0.900, 0.966/0.875, and 0.962/0.880 for the OD and OC on the respective datasets, while significantly reducing inference time.

Anxiety in older individuals is understudied despite its prevalence. Investigating its occurrence can be challenging, yet understanding the factors influencing its recurrence is important for improved management. This study aimed to model the recurrence of anxiety symptomatology in an older population within a five-month timeframe. Data included baseline socio-demographic and general health information for older adults aged 60 years or older with at least mild depressive symptoms. A Bayesian network model explored the relationship between baseline data and recurrent anxiety symptomatology. Model evaluation employed the Area Under the Receiver Operating Characteristic Curve (AUC). The Bayesian model was also compared to three machine learning models. The model achieved an AUC of 0.821 on the test data, using a threshold of 0.367. This result surpassed both the Logistic Regression and XGBoost models, but slightly behind SGDClassifier. Key factors associated with recurrence of anxiety symptomatology were: “Not being able to stop or control worrying”; “Becoming easily annoyed or irritable”; “Trouble relaxing”; and “depressive symptomatology severity”. The model demonstrated generalisation abilities and outperformed some machine learning models while being less complex and more explainable. These findings indicate a prioritised sequence of predictors to identify individuals most likely to experience recurrent anxiety symptomatology.

With the exponential growth of internet usage and mobile devices, the volume of user behavior data has surged, offering significant opportunities and challenges for businesses. This paper explores user behavior analysis in the context of big data, highlighting its pivotal role in driving innovation and strategic decision-making. We begin by discussing the importance of user behavior data and reviewing technological advancements such as machine learning, natural language processing, and data mining that enable deep insights into user behavior. Through a case study on linguistic analysis in livestreaming e-commerce, we demonstrate how text-mining techniques can correlate linguistic characteristics with sales performance, providing actionable marketing insights. We further explore the broader business applications, including personalized recommendations, user portraits, product design, and strategic decision-making. Technical and ethical challenges, such as data integration, privacy concerns, and algorithmic bias, are also addressed. The paper concludes with a discussion on future trends, such as AI and ML advancements, edge computing, IoT proliferation, and ethical AI frameworks, poised to shape user behavior analysis. Our findings underscore the transformative potential of user behavior data in enhancing business operations and user experiences, providing recommendations for future research focused on real-time analytics, privacy-preserving techniques, and ethical implications of data-driven decisions.

The joint extraction of entities and relations is a critical task in information extraction, and its performance directly affects the performance of downstream tasks. However, existing joint extraction models based on deep learning exhibit weak processing capabilities for the phenomenon of multiple pronunciations of one character and multiple characters of one pronunciation when processing Chinese texts, resulting in performance loss. To address these issues, this paper introduces part-of-speech (POS) and pinyin features to aid the model in learning semantic features that are more contextually appropriate. A Chinese Joint Entity and Relation Extraction Model based on Progressive Contrastive Learning and Heterogeneous Feature Fusion is proposed (CJE-PCHF). During model training, an interactive fusion network based on progressive contrastive learning is employed to learn the dependencies between pinyin, POS, and semantic features. This guides the model in heterogeneous feature fusion, capturing higher-order semantic associations between heterogeneous features. On the commonly used DuIE evaluation dataset for joint extraction, this model achieved a significant improvement, with the F1 score increasing by 5.4% compared to the benchmark model CasRel.

The success of deep learning (DL) in offline brain-computer interfaces (BCIs) has not yet translated into efficient online applications. This is due to two limiting factors: the need for large amounts of training data in DL and the fact that current DL solutions are primarily developed for offline decoding. To enable real-time decoding, even across subjects without calibration data, we first introduce a novel method real-time adaptive pooling (RAP) to tune existing offline DL models towards online decoding by modifying the pooling layers. To reduce the amount of training data required, our approach leverages different levels of domain adaptation. We show how different settings enable different adaptation solutions. Our results demonstrate that our approach is both powerful and can be calibration-free, providing a robust and practical solution for real-time BCI applications. These findings pave the way for the development of co-adaptive and highly efficient DL-based BCI systems.

The proliferation of fake news on social media platforms poses significant challenges to information authenticity and public trust. In this study, the impact of various word embedding techniques on the performance of both machine learning and deep learning models for fake news detection is systematically evaluated. The comprehensive TruthSeeker dataset, which spans over a decade of labeled news articles and social media posts, is utilized. Three popular word embedding techniques are analyzed: TF-IDF, Word2Vec, and FastText. These embeddings were tested with multiple machine learning classifiers, including logistic regression, naive Bayes, K-nearest neighbors, support vector machines, multilayer perceptrons, decision trees, and random forests, as well as deep learning models, specifically convolutional neural networks (CNNs). Among machine learning models, the Support Vector Machine (SVM) with TF-IDF achieved the highest accuracy of 99.03%, precision of 98.92%, recall of 99.15%, and F1 score of 99.03%. The Multilayer Perceptron (MLP) also demonstrated strong performance with Word2Vec, reaching an accuracy of 95.24%, precision of 94.95%, recall of 95.50%, and F1 score of 95.07%. For deep learning models, CNNs combined with TF-IDF embeddings showed outstanding results, achieving an accuracy of 98.99%, precision of 98.56%, recall of 99.48%, and F1 score of 99.02%. These results highlight the critical role of word embeddings in enhancing model accuracy and reliability in misinformation detection.

Countless disasters have resulted from climate change, with severe damage to infrastructure and the economy. Their societal impacts require mental health services for the millions affected. To prepare for and respond effectively to these events, it is important to understand people’s emotions and incidents both before and after a disaster strikes. The emotions and other life incidents expressed on Twitter during the 2017 Hurricane Harvey were mined in this case study. Between August 20 and August 30, we collected a dataset of approximately 400,000 public tweets that related to the storm. Using a BERT-based model, we predicted the emotions attached to each tweet. After that, we used various natural language processing techniques to analyze the tweets themselves to understand what kind of ’topics’ were being discussed when people expressed certain ’emotions’. Applying the Latent Dirichlet Allocation (LDA) technique for topic modeling sidesteps the need for manual content analysis and allows us to find meaningful patterns in the data. Results show that there is a significant change of emotions occurred when the hurricane is at most devastating and many of life incidents become pressing demands/concerns from population such as evacuation plan, animal safety, safety update, public policy, helps, flood, and fake news.

The different systems in the nuclear power plants (NPP) are critical and complex which requires a continuous rigorous monitoring for both normal and abnormal conditions. However, due to the nonlinearity of the dynamic behavior of these systems, implementation of artificial intelligence within the NPP components is crucial to enhance the monitoring and the predictability of the key operating parameters trend. On the other hand, lessons learned from large nuclear accident proves that a remote real-time coordination between the different stakeholders involved in the safety of the nuclear reactor is needed. This remote feature can be implemented in the existing power plants by embedding a mobile computing networks within the plant components. This network will send early warning and transmit the NPP data on real-time to different authorities such as the regulatory authority and the decision-making committees, to enhance the interpretability of the nuclear event and deliver a collaborative decision which in return mitigate the risk associated with these events and increase the overall safety of the plant. The integration of AI and mobile computing in NPP would be most of interest in countries where nuclear reactors data are challenged to reach during an accident. For instance, during the meltdown of Chernobyl, the Soviet government covered the accident from the news in an attempt to contain the consequences but there was a spread of radioactive contamination to some other Europe countries. Therefore, in the case of such scenario the implementation of mobile computing and AI would result in different outcomes as the European countries would be notified through the mobile networking and the spread of contamination would be forecasted by the AI algorithms. In this paper, we examine the different modern AI and mobile computing technologies, the different potential application, the associated features and challenges, and future work direction.

Machine learning has gained significant recognition within healthcare organizations as a potent technology for medical image analytics. Among various medical image modalities, Contrast-Enhanced CT (CECT) is widely utilized to acquire additional diagnostic information that can help to better visualize and evaluate certain structures or abnormalities in the human body. Accordingly, there is an increasing demand for machine learning systems for disease analytics using CECT. However, developing such systems presents two significant challenges: high technical complexity and high development effort. This paper presents a software platform that can effectively remedy these challenges. This platform embodies a unified software process for CECT-specific disease diagnosis including oncological diseases, cardiovascular diseases, and gastrointestinal diseases. It integrates multiple types of machine learning models into the disease diagnosis process, where the models can diagnose lesions, malignancies, tumors, medical features of tumors, temporal transitions with a contrast agent, and disease progress. The platform has been implemented in Python using Scikit-learn and TensorFlow libraries. To validate its applicability and reusability, a Hepatocellular Carcinoma (HCC) diagnosis system has been implemented.

The advancement of urban scholarship and the effective addressing of urban environment challenges necessitate the adoption of sophisticated analytical methods. Urban scholars and policymakers need advanced analytical methods to tackle issues like gentrification, housing affordability, and urban sprawl. Predictive models are crucial in the realm of urban sciences, and hyperparameter tuning methods can significantly improve their accuracy and efficiency. Our study compares three such methods — Optuna, Random Search, and Grid Search — using a housing transaction dataset. We find that Optuna is not only 5.58 to 70.50 times faster than the other two methods when applied to Random Forest and Gradient Boosting Machine algorithms, but also achieves lower error values in key evaluation metrics on the test set, such as mean absolute error, mean squared error, mean absolute percentage error and root mean squared error.

Hyperspectral small target detection (HSTD) is a promising pixel-level detection task. However, due to the low contrast and imbalance number between the target and background spatially and high dimensions spectrally, it is a challenging one. To address these issues, this work proposes a representation learning-based graph and generative network for hyperspectral small target detection. The model builds a fusion network through frequency representation for HSTD, where the novel architecture incorporates irregular topological data and the spatial-spectral feature to improve its representation ability. Firstly, a graph convolutional network (GCN) module better models the non-local topological relationship between samples to represent the hyperspectral scene’s underlying data structure. The mini-batch-training pattern of the GCN decreases the high computational cost of building an adjacency matrix for high-dimensional data sets. In parallel, the generative model enhances the differentiation reconstruction and the deep feature representation ability with respect to the target spectral signature. Finally, a fusion module compensates for the extracted different types of HS features and integrates their complementary merits for hyperspectral data interpretation while increasing detection and background suppression capabilities. Experiments on different hyperspectral data sets demonstrate the advantages of the proposed architecture.

This paper comprehensively explores the integration of cloud computing and advanced recommendation systems, emphasizing their pivotal roles in enhancing user experiences and operational efficiencies across digital platforms. It reviews the evolution of recommendation algorithms, highlighting their application in diverse domains such as e-commerce and media. The study evaluates the performance of advanced models like UniLLMRec against traditional counterparts using datasets from news and e-commerce domains. Additionally, the paper discusses the infrastructure architecture of cloud computing, demonstrating its capability to support scalable and efficient data processing. Through experimental insights and methodology, the research underscores the transformative impact of cloud technologies on optimizing recommendation system performance, thereby advancing digital engagement and competitiveness.

Logistic Regression and Random Forest are used to identify risk factors that influence traffic accident fatalities in the United Kingdom. The mean decrease accuracy was used to measure variable importance. The speed limit, police attendance and quarter had an increasing influence on accident fatalities. They had a mean decrease of 102.1669, 221.5322, and 120.894 respectively. The speed limit, had a parameter estimate of 0.0046902 and a standard deviation of 0.0004875. Light Conditions: Night had a parameter estimate of 1.2657635 and a standard deviation of 0.0118409. Road Type Round About had a parameter estimate of -0.4055796 and a standard deviation of 0.0210848. Police Attendance classified as Yes had a parameter of 0.8546232 and a standard deviation of 0.0151043. The best predictors were speed limit, police attendance and quarter since they had p values that were less than 0.05. The findings of the study indicated that logistic Regression had a higher accuracy rate 79.85% as compared to 64.00% for Random Forest. A split test was used and a standard deviation of 0.0010486 was obtained for the Logistic Regression model.

Credit card payment platforms are increasingly used for e-commerce activities. Credit cards are payment cards issued to individuals so they can purchase products and services based on their accumulated debt. Since credit cards are becoming very popular and widely used, the number of related fraud cases is likewise increasing. Whereas, there are large numbers of credit card transactions daily; thus, it becomes complex to differentiate between fraudulent and non-fraudulent transactions. Accordingly, different machine learning (ML) tools have been deployed in the literature to filter credit card transactions to prevent cardholders and financial institutions from losing money. In this article, the ensemble hidden Markov model (EHMM) approach is deployed to predict credit card fraud. EHMM is a popular and flexible ML tool that can easily model randomly changing datasets. Yet, the performance of the EHMM solely depends on the adopted feature extraction technique. The more reliable the feature extraction technique, the better the performance of the EHMM. In this vein, this article analyses two feature extraction techniques that can be combined with the EHMM for the prediction of credit card fraud. First, the principal component analysis (PCA) feature extraction techniques are combined with the EHMM to effectively predict credit card fraud. Yet, the PCA increases the computational time complexity of the EHMM. Therefore, this article adapts some statistical features to reduce the computational time complexity imposed on the EHMM. These robust but simple statistical features termed MRE; Mean, Relative Amplitude, and Entropy, are merged to form a feature vector that can be combined with the EHMM to effectively predict credit card frauds. The performance of the PCA-EHMM and MRE-EHMM are evaluated using the credit card transactions dataset of European cardholders gathered within two days in September 2013. The results were documented using different performance metrics such as recall/sensitivity, specificity, precision, and F1-score. The findings of this article will offer solutions to the loss experienced by cardholders and financial institutions as a result of online banking which requires credit card details.

Generative Adversarial Networks (GANs) are powerful models for generative tasks but face significant challenges such as instability, mode collapse, and inefficient convergence. This paper introduces a theoretical framework leveraging Fermi-Dirac statistics, Ruppeiner geometry, phase transition dynamics, and graph theory to address these issues. The proposed theorems provide a robust mathematical foundation for enhancing GAN stability, diversity, and convergence. Experimental validation on the MNIST dataset demonstrates that the application of these theorems results in improved Inception Scores, reduced Fr´echet Inception Distances, and stable training dynamics, as indicated by Ruppeiner curvature analysis. The findings suggest that these theoretical insights offer a comprehensive solution to the persistent challenges in GAN training, paving the way for more reliable and effective generative models.

Renewable energy especially insights into solar energy are crucial in the energy business cycle. In this paper, an impact study of meteorological variables on solar energy production is conducted. The paper discusses the pipeline of data analysis specific to the case study and exploits open-source machine-learning algorithms to build predictive models. RandomForestRegressor and XGBoosting regressors are used to build the predictive model. The application code is integrated into an analytic software service (Power BI) and tested on Solar energy Open Data of Saudi Arabia. The solar energy production is found to be high between 800-1000 barometric pressure locations. Global Horizontal Irradiation is found to increase when the atmospheric temperature is more than 34∘ C. The impact of relative humidity is that the produced energy is high at lower humidity points. The wind speed effect is that the produced energy is significant between 2 to 4 m/s and reaches its maximum at 2.4 m/s. The predictive models produce nearly real-time predictions with Random Forest with R²: 0.909. The paper provides Global Horizontal Irradiation estimation equations using meteorological coefficients generated by the trained regressor. The paper presents implementation methodologies and records various instances of test results. It highlights the efficacy of machine learning in predictive modeling for renewable energy applications.

In an era where the proliferation of large language models blurs the lines between human and machine-generated content, discerning text authenticity is paramount. This study investigates transformer-based language models—BERT, RoBERTa, and DistilBERT—in distinguishing human-written from machine-generated text. By leveraging a comprehensive corpus, including human-written text from sources such as Wikipedia, WikiHow, various news articles in different languages, and texts generated by OpenAI's GPT-2, we conduct rigorous comparative experiments. Our findings highlight the superior effectiveness of ensemble learning models over single classifiers in this critical task. This research underscores the versatility and efficacy of transformer-based methodologies for a wide range of natural language processing applications, significantly advancing text authenticity detection systems. The results demonstrate competitive performance, with the transformer-based method achieving an F-score score of 0.83 with RoBERTa-large (monolingual) and 0.70 with DistilBERT-base-uncased (multilingual).

This study presents the Distilled MixUp Squeeze ResNet (DMSResNet), an optimized variant of the Residual Network architecture designed for high accuracy and computational efficiency. By integrating Squeeze-and-Excitation blocks, MixUp data augmentation, and knowledge distillation, it achieves competitive performance within strict parameter constraints and limited training resources. Evaluated on the CIFAR-10 dataset, the model attains a 96.56% accuracy with only 4.3 million total parameters. The results demonstrate that optimization of established architectures can yield performance comparable to newer models, especially in resource-limited scenarios. This work contributes to the development of efficient deep learning models for applications in constrained computational environments.

Financial market data analysts are increasingly turning towards machine learning techniques for data analysis and making decisions. This paper explores how knowledge graphs, essentially a network of real entities and their relationships, can be used to improve collaborative financial market data analysis and make it more efficient and user friendly. This paper reviews the current state of financial market data analysis, the challenges in deploying machine learning models in industrial environments, and the concepts such as ML-Ops and knowledge graphs. A software architecture comprising various roles, layers and services such as Data Engineers, Data Analysts, public users, APIs, workflows, analytics libraries, UX layer, Business Layer, etc. is introduced and described in detail. Then, it discusses how knowledge graphs can be used to enhance collaborative financial market data analysis. Finally, a case study is presented to demonstrate the usage of the whole system, within the context of financial market data analytics of equities.

Which machine-learning model is the best at winning the prisoner’s dilemma? Which models create the best cumulative outcomes? Is there a model that perfectly captures both winning and cumulative points? These are questions generated from a simple 2 x 2 payoff matrix of the prisoner’s dilemma. Imagine you and your partner in crime are caught and sent independently into questioning. You can either collaborate or defect, but you don’t know what your partner will do. If you both collaborate, cumulatively, you’ll each get one year in jail. If you defect and your partner collaborates, you’ll serve no time and they will serve 10 years, and vice versa. If you both defect, you’ll both serve 5 years in jail. Placing AIs against each other to play just one round doesn’t reveal much about their code, strategy, and end goal. So the Reinforcement Learning, Pattern Learning, Tit For Tat, and other models were put up against each other in a 100 round game where their behavior, convergence, and learning were analyzed to reveal the most effective ways strategies to beat the prisoner’s dilemma. All code and data is open sourced here.

Named Entity Recognition (NER) is a fundamental and pivotal stage in the development of various knowledge-based support systems, including knowledge retrieval and question-answering systems. In the domain of pig diseases, Chinese NER models encounter several challenges such as the scarcity of annotated data, domain-specific vocabulary, diverse entity categories, and ambiguous entity boundaries. To address these challenges, we propose PDCNER, a Pig Disease Chinese Named Entity Recognition method leveraging lexicon-enhanced BERT and contrastive learning. Firstly, we construct a domain-specific lexicon and pre-train word embeddings in the pig disease domain. Secondly, we integrate lexicon information of pig diseases into the lower layers of BERT using a Lexicon Adapter layer, which employs char-word pair sequences. Thirdly, to enhance feature representation, we propose a lexicon-enhanced contrastive loss layer on top of BERT. Finally, a Conditional Random Field (CRF) layer is employed as the model's decoder. Experimental results show that our proposed model demonstrates superior performance over several mainstream models, achieving a precision of 87.76%, a recall of 86.97%, and an F1-score of 87.36%. The proposed model outperforms BERT-BiLSTM-CRF and LEBERT by 14.05% and 6.8%, respectively, with only 10% of the samples available, showcasing its robustness in data scarcity scenarios. Furthermore, the model exhibits generalizability across publicly available datasets. Our work provides reliable technical support for the information extraction of pig diseases in Chinese and can be easily extended to other domains, thereby facilitating seamless adaptation for named entity identification across diverse contexts. The codes have been open-sourced at https://github.com/tufeifei923/pdcner.

Large language models (LLMs) have recently garnered significant attention due to their remarkable ability to assimilate vast amounts of information and effectively process natural language. In healthcare, natural language constitutes a substantial portion of medical data, rendering LLMs highly promising for various healthcare applications. This study seeks to explore the potential of LLMs in healthcare and clinical decision support (CDS), following PRISMA guidelines for reviews. The analysis encompasses 44 LLMs, each influenced by several factors impacting their performance. Notably, the datasets utilized for pretraining and finetuning processes play a crucial role in determining the model’s domain specificity. Furthermore, distinct model architectures are tailored for specific tasks, while prompting strategies are frequently employed to refine and enhance the model’s performance. LLMs exhibit considerable promise for a wide array of healthcare applications. For instance, LLMs possess the potential to handle and analyze medical information efficiently, facilitate contextual understanding among clinicians and patients, as well as automate the documentation of clinical notes and reports. Presently, however, their implementation within the field remains limited. Notable improvements have been witnessed in the performance of current healthcare-oriented LLMs, with some achieving expert-level competence in medical question-answering (MQA). However, these LLMs face prominent challenges, encompassing ethical concerns, issues related to accountability, and a lack of appropriate regulations. Nevertheless, this study reveals numerous promising applications in healthcare where LLMs could significantly augment the efficiency, accessibility, and manageability of healthcare delivery. Addressing the challenges LLMs encounter is essential for their seamless integration into practical healthcare applications. As a relatively new technology, the development of LLMs is still in its early stages, but their potential is evident through this study. Consequently, fostering collaboration among healthcare professionals, developers, regulators, and other stakeholders is imperative to cultivate dependable LLMs that align with the demands of the healthcare sector.

This study introduces a data obfuscation technique, leveraging the exponential map associated with the generators of Lie groups. Originating from quantum machine learning frameworks, our method illustrates the practical application of quantum mechanics principles in data processing. Specifically, it employs the exponential map of a generator algebra to introduce controlled noise into the data, achieving obfuscated data while preserving its utility for machine learning tasks. This strategy is shown to safeguard privacy in sensitive datasets, such as discussed medical records, and to enhance dataset volume and diversity through augmentation. Our empirical analysis, benchmarked against standard machine learning approaches, demonstrates that our method can maintain or even improve the predictive accuracy of the original data. This research highlights the potential of Lie group theory for advancing data privacy in medicine, marking a significant contribution to machine learning methodologies by offering the dual benefits of data obfuscation and enrichment. Through this synthesis of algebraic structures and machine learning, we propose new pathways for the secure and effective use of data in sensitive areas.

This article aims to analyze household energy data to predict electricity self-sufficiency and identify the key features that impact it. For this purpose, we use SHAP (Shapley Additive Explanations), LightGBM (Light Gradient Boosting Machine), and a correlation heatmap-based framework to analyze 12 months of energy and questionnaire survey data collected from over 200 smart houses in Kitakyushu, Japan. We use the SHAP summary plot to identify the impact-wise order of key features influencing the electricity self-sufficiency rate (ESSR). Using SHAP, we demonstrated that key features are; housing types, average monthly electricity bill, floor heating, electric capacity, number of washing and drying machines, fee plans, occupation of household head, total floor area, average monthly gas bill, cooking equipment, etc. Furthermore, we adopted the LightGBM classifier with $\ell^1$ regularization to extract the most significant features and predict the electricity self-sufficiency rate of households. This LightGBM-based model can also predict the electricity self-sufficiency rate of households that did not participate in the questionnaire survey. A heat map is also used to analyze the correlation among household variables. The findings discussed in this article offer valuable insights for energy policymakers to achieve the targets of energy self-sufficient houses.

The increasing complexity and volume of job listings in the digital era necessitate advanced recommendation systems to match job seekers with suitable opportunities. This paper presents a hybrid job recommendation system that integrates Collaborative Filtering (CF) and Content-Based Filtering (CBF) to enhance recommendation accuracy and user satisfaction. The CF component leverages historical user interactions and similarities between users to suggest jobs, while the CBF component analyzes job descriptions and user profiles to provide personalized recommendations based on individual preferences and qualifications. By combining these two approaches, the proposed system mitigates the limitations of each method and offers a more comprehensive and effective solution. Experimental results on a real-world dataset demonstrate the system's improved performance in terms of precision, recall, and F1-score compared to traditional recommendation models. This study contributes to the field of job recommendation systems by providing a robust framework that effectively bridges the gap between user needs and job market opportunities.

Link prediction is a fundamental problem in graphs. Currently, SubGraph Representation Learning (SGRL) methods provide state-of-the-art solutions for link prediction by transforming the task into a graph classification problem. However, existing SGRL solutions suffer from high computational costs and lack scalability. In this paper, we propose a novel SGRL framework called Augmented Feature Diffusion on Sparsely Sampled Subgraph (AFD3S). The AFD3S first uses a conditional variational autoencoder to augment the local features of the input graph, effectively improving the expressive ability of downstream Graph Neural Networks. Then, based on a random walk strategy, sparsely sampled subgraphs are obtained from the target node pairs, reducing computational and storage overhead. Graph diffusion is then performed on the sampled subgraph to achieve specific weighting. Finally, the diffusion matrix of the subgraph and its augmented feature matrix are used for feature diffusion to obtain operator-level node representations as inputs for the SGRL-based link prediction. Feature diffusion effectively simulates the message-passing process, simplifying subgraph representation learning, thus accelerating the training and inference speed of subgraph learning. Our proposed AFD3S achieves optimal prediction performance on several benchmark datasets, with significantly reduced storage and computational costs.

In the last few years, the agricultural field has undergone a digital transformation, incorporating artificial intelligence systems to make good employment of the growing volume of data from various sources and derive value from it. Within artificial intelligence, machine learning is a powerful tool for confronting the numerous challenges of developing knowledge-based farming systems. This study aims to comprehensively review the current scientific literature from 2017 to 2023, emphasizing machine learning in agriculture, especially viticulture, to detect and predict grape infections. Most of these studies (88%) were conducted within the last five years. A variety of machine learning algorithms were used, with those belonging to the neural network (CNN) standing out having the best results most of the time. Out of the list of diseases, the ones most researched were Grapevine Yellow, Flavescence Dorée, Esca, Downy mildew, Leafroll, Pierce’s, and Root Rot. Also, some other fields were studied which are water management, plant deficiencies, and classification. Because of the difficulty of the topic, it was collected all datasets that were available about grapevines, and it is described each dataset with the type of data (e.g., statistical, images, type of images), along with the number of images where it was mentioned. This work provides a unique source of information for a general audience comprising AI researchers, agricultural scientists, wine grape growers, and policymakers. Among others, its outcomes could be effective at curbing diseases in viticulture, which in turn will drive sustainable gains and boost success. Additionally, it could help build resilience in related farming industries such as winemaking.

Artificial General Intelligence (AGI) represents a pivotal advancement in AI with far- reaching implications across technological, ethical, and societal domains. This paper addresses the following: (1) an in-depth assessment of AGI’s transformative potential across different sectors and its multifaceted implications, including significant financial impacts like workforce disruption, income inequality, productivity gains, and potential systemic risks; (2) an examination of critical ethical considerations, including transparency and accountability, complex ethical dilemmas and societal impact; (3) a detailed analysis of privacy, legal and policy implications, particularly in intellectual property and liability, and (4) a proposed governance framework to ensure responsible AGI development and deployment. Additionally, the paper explores and addresses AGI's political implications, including national security and potential misuse. By analyzing and considering computer science, philosophy, economics, and policy perspectives, we offer a multidisciplinary view of AGI’s challenges and opportunities, advocating for proactive measures to align AGI development with human values and societal interests.

Large language models (LLMs) have shown impressive performance and versatility. However, their billions of parameters and high computational costs hinder the development of personalized and privacy-preserving AI assistants operating locally on user devices. In this work, we explored the potential of pruning LLMs to create lightweight models suitable for user devices, using the moderate-sized Llama-2 7B model as an example. By adopting a simple yet effective pruning method, we found that up to 60% of the weights in the Llama-2 7B model could be pruned without significantly impairing its language modeling capabilities. Furthermore, despite occasional factual inaccuracies, the pruned model at the sparsity limit generated fluent and helpful answers to daily queries, demonstrating its feasibility of on-device AI assistants. These inaccuracies might originate from forgetting or hallucination due to pruning. We proposed a simple protocol to distinguish between the two mechanisms, as well as future directions to improve the pruned models for local AI assistants.

Machine reading comprehension is a crucial area of research in natural language processing, aiming to enable machines to read and understand text as humans do, and to answer questions related to the text content. The pre-trained language model, represented by BERT, has achieved superior results compared to traditional models in many NLP tasks, leading to the rise of the pre-trained paradigm in the field of natural language processing. This paper addresses the issue that the pre-trained language model lacks the ability to extract long-distance semantic relations and make efficient use of linguistic features. Firstly, the recent developments of the pre-trained language model are described. Then, two feature maps are used to intuitively express the structured long-distance semantic correlation features, and the traditional sequence structure features are integrated. The influence of different feature map construction methods on machine reading comprehension is compared. Finally, the application of the pre-trained language model in machine reading comprehension is summarized and prospected. By fusing graph structure features, the model can learn more abundant language knowledge, thus further improving its reasoning ability.

The integration of Artificial Intelligence (AI) in predicting and managing risks for coastal port developments significantly enhances planning and operational efficiency. This paper presents a theoretical framework utilizing AI to predict future scenarios based on geological, hydrodynamic, and oceanographic factors. Focusing on the effective implementation of AI in ports, with a specific emphasis on Bulgarian Black Sea coastal areas such as Nesebur, Sarafovo, Sozopol, St. Anastasia Island, and the region of Karantinata in Varna, the study explores AI applications for site selection, risk assessment, and operational optimization to ensure resilient, efficient, and sustainable port infrastructure.

Classification of pre-processed fMRI dataset using the functional connectivity(FC) based features is considered a challenging task due to high dimensional FC features set and smaller dataset size. To tackle this specific FC high dimensional features set and a smaller dataset size , We have proposed here a GAN based dataset augmenter to firstly train the GAN on the NYU Connectivity Features dataset and used the trained GAN to generate synthetic features per category and after getting sufficient number of features per category , A multi-head attention mechanism was used as a head for the classification . We name our proposed approach as "ASD-GANNet" which is End-To-End and does not require hand-crafted features as the multi-head attention mechanism focuses on the features that are more relevant . Moreover , We compared our results with the 06 available state-of-the-art techniques from the literature and Our proposed approach results using "NYU" site as a training set for generating GAN based synthetic dataset are promising. We achieved an overall 10-fold cross validation based accuracy of 82% ,sensitivity 82% and specificity 81% outperforming the available state-of-the art approaches. Sitewise comparison of our proposed approach also outperformed the the available state-of-the-art as out of the 17 sites our proposed approach has better results in the 10 sites.

Objectives. The aim of this study was to demonstrate the ability to visualize the results of the Scilit platform's bibliometric data analysis on the topic "AI & Machine Learning" to identify publications reflecting specific issues of the topic. Data source. Bibliometric records exported from the Scilit platform on the topic "AI & Machine Learning" for the years 2021–2023 were used. For each year, 6,000 records were downloaded in CSV and RIS format. Programs and utilities used. VOSviewer, Scimago Graphica, Inkscape, FP-growth utility, GSDMM algorithm. Used services: Elicit, QuillBot, Litmaps. Results. It has been shown that bibliometric data from the open access abstract database Scilit can serve as a quality alternative to subscription-only databases. Data exported from the Scilit platform require preprocessing to make them available in a format that can be processed by programs such as VOSviewer and Scimago Graphica. The use of GSDMM and FP-growth algorithms is effective for structuring bibliometric data for further visualization. The Scimago Graphica software provides wide possibilities for building compound diagrams, in particular, for representing the network of keywords in such important coordinates for bibliometric analysis as average year of publication and average normalized citation, as well as for building an alluvial diagram of co-occurrence of more than two keywords. The possibility of using such services as elicit.com, quillbot.com and app.litmaps.com to accelerate the selection of publications on the topic under study is shown.

This paper systematically discusses the application of reinforcement learning in automatic driving system. Reinforcement learning frameworks show significant advantages in optimizing decision making, predictive perception, path planning, and controller design, exceeding the limitations of traditional supervised learning methods. The paper highlights the critical role of components such as scene understanding, positioning, and map making in autonomous driving systems, which provide reliable environmental awareness through deep learning and sensor fusion technologies to support intelligent decision-making in complex urban environments. In addition, the paper discusses innovative approaches to safety reinforcement learning to reduce risk in autonomous driving and ensure that systems adhere strictly to safety constraints while maximizing expected rewards. These findings provide an important theoretical and practical basis for further improving algorithm robustness, managing multi-agent interactions, and integrating ethical considerations in the future.

Disaster issue consultation is crucial for effective emergency management, requiring timely and accurate information to support decision-making processes. This paper explores the integration of large language models (LLMs) with advanced prompt engineering to enhance the performance of AI systems in disaster management. We propose a novel multi-round prompt engineering approach that guides LLMs to iteratively refine their responses based on evolving disaster scenarios. Our method leverages diverse disaster-related datasets and employs a combination of accuracy metrics and a customized GPT-4 scoring system for evaluation. We conduct comprehensive experiments across multiple LLM platforms, including Qwen2, ChatGPT, Claude, and GPT-4, comparing our approach with baseline models and the Tree of Thoughts (ToT) method. The results demonstrate significant improvements in both accuracy and response quality, highlighting the effectiveness of our method in providing contextually relevant and actionable insights. Additional analysis confirms the robustness and generalizability of our approach across various disaster scenarios, establishing it as a valuable tool for enhancing disaster management practices.

With the continuous advancement of digital transformation, digital payments are playing an increasingly important role in the financial industry. This study aims to utilize machine learning models to predict and analyze digital payment behavior. Initially, the background and significance of digital payments in the financial sector are introduced. Subsequently, the current status and trends of traditional digital payment distribution are reviewed, alongside related work on digital payment behavior prediction. Methodologically, principles and applications of machine learning models such as logistic regression, decision trees, and random forests are elaborated, along with experimental design and data preprocessing methods. The experimental results and discussion section illustrates the performance of each model in digital payment prediction and explores their impact on credit decisions. This exploration equips financial institutions with more effective user behavior analysis and risk management tools, thereby fostering future development and application of digital payment technologies.

Since the beginning of the internet era, there has been an explosion of growth in structured data (such as numbers, symbols, and labels) as well as unstructured data (including images, videos, and text). Efficient and accurate mixed query of these two types of data is a key technology to achieve high-quality information retrieval, and it is also a major challenge that needs to be solved urgently in the industry. In this study, we employ an advanced Transformer model that combines strategies and fine-tuning techniques for multi-task learning. Specifically, the model is first pre-trained on a large-scale, general-purpose dataset to learn different types of data representations and basic language comprehension skills. After that, we fine-tuned the parameters of the model to better suit these specific data processing tasks for specific application scenarios, such as image annotation, video content analysis, and structured data query. At the heart of the model is the self-attention mechanism, which allows the model to automatically emphasize the important parts and ignore irrelevant information when processing the input data. In addition, we have introduced task-specific adaptation layers that are designed to add additional processing power to the original Transformer architecture, such as a semantic analysis layer for unstructured text data and a relational extraction layer for structured data. This combination of general pre-training and task-specific fine-tuning allows the model to flexibly process and integrate information from different data sources, improving processing efficiency and accuracy. Experimental results show that the model performs well in a variety of data processing tasks, significantly improves the accuracy and efficiency of information retrieval, and verifies the strong potential and adaptability of large language models in processing mixed data types.

In a trip planning service, travelers first set a starting point, a destination, and a sequential list of specific points of interest types (e.g., museums, restaurants, and parks). Based on this information, the service searches the spatial database to customize the best travel itinerary for the tourist. However, in previous studies, planners only considered the time factor when designing the optimal route, and failed to fully consider the quality of each point of interest. In this study, we specifically leveraged the capabilities of large language model to parse and respond to complex travel-related user queries. To apply large language model to route planning, we fine-tuned the model to understand geotagging and user travel preferences. We have introduced a novel graph search algorithm combined with large language model output, which optimizes the route search process to provide optimal travel recommendations by taking into account various factors such as distance length, budget constraints, and popularity of tourist attractions. In addition, we have integrated real-time traffic data and historical travel data to further improve the prediction accuracy and application usefulness of the model. In the experimental validation phase, we designed a series of benchmarks to compare the performance of the system with traditional algorithms and other machine learning-based route planning methods. Experimental results show that our model has a significant improvement compared with traditional methods in improving the speed and accuracy of path selection.

Credit business is a vital part of the bank's core business, which has an extremely important impact on the bank's income and development. In the operation of credit business, credit risk assessment is particularly crucial, and accurate risk assessment can minimize risks while maximizing the bank's returns. We propose a method to optimize credit risk assessment using machine learning techniques. In this work, we employ a random forest machine learning model to process and analyze large amounts of loan application data. By using correlation analysis, information enrichment, etc., the characteristics that have the most impact on credit risk assessment are screened. Subsequently, the model was constructed using a random forest algorithm. Random forests improve the generalization ability and accuracy of the model by building multiple decision trees and introducing randomness between these trees. In the experimental analysis part, we compare the performance of various models on the German credit dataset, and the results show that the deep learning model outperforms the traditional machine learning model in most indicators, verifying the effectiveness of our method.

The use of artificial intelligence algorithms (AI) has gained importance for dental applications in recent years. Analyzing using IA information from different sensor data such as images or X-ray radiographs can help to improve medical decisions and to achieve early diagnosis of different dental pathologies. In particular, the use of deep learning (DL) techniques based on convolutional neural networks (CNNs) has obtained promising results in dental applications based on images, in which approaches based on classification, detection, and segmentation are being studied with growing interest. However, there are still several challenges to be tackled, such as the data quality and quantity, the variability among categories, and the analysis of the possible bias and variance associated with each dataset distribution. In this work, we compare three different DL object detection models, which are Faster R-CNN, YOLO V2, and SSD. Each object detection architecture was trained, calibrated, validated, and tested with three different feature extraction CNNs which are ResNet-18, ResNet-50, and ResNet-101, which were the networks that best fit our dataset distribution. Based on such detection networks, we detect four different categories of angles in third molars using X-ray panoramic radiographs by using Winter's classification criterion. This criterion characterizes the third molar's position regarding the second molar's longitudinal axis. The detected categories for the third molars are distoangular, vertical, mesioangular, and horizontal. For training, we used a total of 644 panoramic X-ray images. The results obtained in the testing dataset reached up to 99% mean average accuracy performance, demonstrating the YOLOV2 obtained higher effectiveness in solving the third molar angle detection problem. These results demonstrate that the use of CNNs for object detection in panoramic X-rays represents a promising solution in dental applications.

This study examines the feasibility of using event-related potentials (ERPs) obtained from electroencephalographic (EEG) recordings as biomarkers for long-term memory item classification. Previous studies have identified old/new effects in memory paradigms associated with explicit long-term memory and familiarity. Recent advancements in convolutional neural networks (CNNs) have enabled the classification of ERP trials under different conditions and identification of features related to neural processes at the single-trial level. We employed this approach to compare three CNN models with distinct architectures using experimental data. Participants (N = 25) performed an association memory task while recording ERPs that were used for training and validation of the CNN models. The EEGNET-based model achieved the most reliable performance in terms of precision, recall, and specificity compared with the shallow and deep convolutional approaches. The classification accuracy of this model reached 70% for known items and 62% for unknown items. Good overall accuracy requires a trade-off between recall and specificity and depends on the architecture of the model and the dataset size. These results suggest the possibility of integrating ERP and CNN into online learning tools and identifying the underlying processes related to long-term memorization.

The world health organisation (WHO) revealed approximately 280 million people in the world suffer from depression. Yet, existing studies on early-stage depression detection using machine learning (ML) techniques are limited. Prior studies have applied a single stand-alone algorithm which are unable to deal with data complexities, prone to overfitting and limited in generalisation. To this end, our paper examined the performance of several ML algorithms for early-stage depression detection using two benchmark social media datasets (D1 and D2). More specifically, we incorporated sentiment indicator to improve our model performance. Our experimental results showed that sentence bidirectional encoder representations from transformers (SBERT) numerical vectors fitted into stacking ensemble model achieved comparable F1 scores of 69% in dataset (D1) and 76% in dataset (D2). Our findings suggest that utilising sentiment indicators as additional feature for depression detection yields an improved model performance and thus, we recommend the development of depressive term corpus for future work.

Session-based recommendations which aim to predict subsequent user-item interaction based on historical user behavior during anonymous sessions can be challenging tasks. Two main challenges need to be addressed and improved: (1) How to analyze these sessions to accurately and completely capture users’ preferences, and (2) how to identify and eliminate any interference caused by noisy behavior. Existing methods have not adequately addressed these issues since they either neglected the valuable insight that can be gained from analyzing consecutive groups of items or failed to take these noisy data in sessions seriously and handled them properly, which can jointly impede recommendation systems from capturing users’ real intentions. To address these two problems, we designed a multi-order semantic denoising (MSD) model for session-based recommendations. Specifically, we grouped items of different lengths as varying multi-order semantic units to mine the user’s primary intention from multiple dimensions. Meanwhile, a novel denoising network was designed to alleviate the interference of noisy behavior and provide a more precise session representation. The result of extensive experiments on three real-world datasets demonstrated that the proposed MSD model exhibited improved performance compared with existing state-of-the-art methods in session-based recommendation.

In genomics analysis, with the development of sequencing technology and the rapid growth of data volume, how to effectively identify and analyze important information in genomic data from massive data has become an important challenge. Identify key patterns and features in genomic data through advanced community testing methods. First, genomic data often contains a lot of noise and redundant information that needs to be processed through steps such as normalization, filtering, and dimensionality reduction. Normalization makes data at different scales comparable, filtering out data points with low quality or missing values, and dimensionality reduction reduces data dimensions through principal component analysis (PCA) and retains the main information. The gene co-expression network establishes the connection relationship between genes by calculating the expression similarity matrix between genes, and applies the Louvain algorithm to maximize the modularity of the network, aggregate nodes layer by layer, form a community structure, and identify the complex relationship and potential structure between genes. Finally, we utilize the cancer dataset to evaluate the proposed model. From our extensive experimental results, we can observe that Louvain's algorithm indicates outperformances and accuracy in the recognition and analysis of genomic data.

Predicting heart disease is crucial for early diagnosis and intervention, significantly improving patient outcomes and reducing mortality rates. This study compares various machine learning models, including Logistic Regression, Random Forest, Gradient Boosting, XGBoost, and Long Short-Term Memory (LSTM) networks, using the Cleveland Heart Disease dataset. Comprehensive preprocessing steps were undertaken, such as handling missing values, converting categorical variables to numeric forms, and binarizing the target variable for binary classification. Each model was rigorously evaluated using performance metrics, including accuracy, precision, recall, F1-score, and the area under the receiver operating characteristic curve (AUC-ROC). SHapley Additive exPlanations (SHAP) values were employed to provide insights into feature importance, ensuring model transparency and interpretability. The results indicate that XGBoost outperformed all other models, achieving an accuracy of 90% and an AUC-ROC of 0.94, demonstrating its superior ability to capture complex patterns in the data through advanced optimization techniques and regularization. This study highlights the significant potential of advanced machine learning techniques, particularly ensemble methods like Gradient Boosting and XGBoost, in enhancing heart disease prediction. These models offer higher accuracy and valuable interpretability, making them practical tools for early diagnosis in clinical settings. Future research should focus on integrating these models into healthcare systems and exploring hybrid approaches to further improve predictive performance and clinical applicability.

Image restoration plays a vital role in image processing and pattern recognition in producing better recognition accuracy. It was observed that, most of the traditional restoration algorithms were meant for removing a single type of noise and efforts have been made towards the use of partial differential equation model (PDE), to restore different types of images such as super-resolution, bleeding through and deblurring. This research work deployed hybridized shock filter, shock diffusion of higher order PDE to remove degraded images that have been affected by atmospheric turbulence and water smear images. Degraded images of different types were acquired from various sources. The degraded images were subjected to shock filter algorithm, shock diffusion of higher order PDE. The shock filter was combined with PDE, likewise the shock diffusion was also combined with PDE, and the three algorithms were also combined to test the efficiency of the hybridized method on different degraded images. The result was validated using Mean Square Error, Structural Similarity Index Measure (SSIM), Peak-to-Noise-Signal -Ratio (PNSR) and universal quality index (UQI). From the result, it was observed that the combination of many image restoration algorithms does not give better results whereas, the combination of two of the algorithms has really proved effective.

Background: Despite recent advancements, medical technology has not yet reached its peak. Precision medicine is growing rapidly, thanks to machine learning breakthroughs powered by increased computational capabilities. This article explores a deep-learning application for computer-aided diagnosis in dermatology. Methods: Using a custom model based on EfficientNetB3 and deep learning, we propose an approach for skin lesion classification that offers superior results with smaller, cheaper, and faster inference times compared to other models. The skin images dataset used for this research includes 8,222 files selected from the authors' collection and the ISIC2019 archive, covering six dermatological conditions. Results: The model achieved 95.4% validation accuracy on four categories—melanoma, basal cell carcinoma, benign keratosis-like lesions, and melanocytic nevi—using an average of 1,600 images per category. Adding two categories with fewer images (about 700 each)—squamous cell carcinoma and actinic keratoses—reduced the validation accuracy to 88.8%. The model maintained accuracy on new clinical test images taken under the same conditions as the training dataset. Conclusions: The custom model demonstrated excellent performance on the diverse skin lesions dataset, with significant potential for further enhancements.

Classifiers face a myriad of challenges in today’s data-driven world, ranging from overfitting and high computational costs to low accuracy, imbalanced training datasets, and the notorious black box effect. Furthermore, many traditional classifiers struggle with the robust handling of noisy and missing feature values. In response to these hurdles, we present classification methods that leverage the power of feature partitioning learning and outranking measures. Our classification algorithms offer an innovative approach, eliminating the need for prior domain knowledge by automatically discerning feature intervals directly from the data. These intervals capture essential patterns and characteristics within the dataset, empowering our classifiers with newfound adaptability and insight. In addition, we employ outranking measures to mitigate the influence of noise and uncertainty in the data. Through pairwise comparisons of alternatives on each feature, we enhance the robustness and reliability of our classification outcomes. The developed classifiers are empirically evaluated on several data sets from UCI repository and are compared with well-known classifiers including k Nearest Neighbors (K-NN), Support Vector Machine (SVM), Random Forest (RF), Neural Network (NN), Naive Bayesian (NB) and Nearest Centroid (NC). The experiments result demonstrate that the classifiers based on feature interval learning and outranking approaches are robust to imbalanced data and to irrelevant features and achieve comparably and even better performances than the well-known classifiers in some cases. Moreover, our proposed classifiers produce more explainable models whilst preserving high predictive performance levels.

Deep learning neural networks require an immense amount of computation, especially in the training phase of the network when networks with multiple layers of intermediate neurons need to be built. In this paper, we will focus on the PSO algorithm with the aim of significantly accelerating the DLNN training phase by taking advantage of the GPGPU architecture and the Apache Spark analytics engine for large-scale data processing tasks. PSO is a bio-inspired stochastic optimization method whose goal is to iteratively improve the solution to a (usually complex) problem by attempting to approximate a given objective. However, parallelizing an efficient PSO is not a straightforward process due to the complexity of the computations performed on the swarm of particles and the iterative execution of the algorithm until a solution close to the objective with minimal error is achieved. In the present work, two parallelizations of the PSO algorithm have been implemented , both designed for a distributed execution environment. The synchronous parallel PSO implementation ensures consistency at the cost of potential idle time due to global synchronization, while the asynchronous parallel PSO approach improves execution time by reducing the need for global synchronization, making it more suitable for large datasets and distributed environments such as Apache Spark. Both variants of PSO have been implemented to distribute the computational load supported by this algorithm –due to the costly fitness evaluation and updating of particle positions– across the different Spark cluster executor nodes to effectively achieve coarse-grained parallelism, resulting in a significant performance increase over current sequential variants of PSO.

The scientific fields of complexity, artificial life (ALife), and artificial intelligence (A.I.) share commonalities: historic, conceptual, methodological, and philosophical. Although their origins trace back to the 1940s birth of cybernetics, they were only able to properly develop as modern information technology became available. In this perspective, I offer a personal (and thus biased) account of the expectations and limitations of these fields, some of which have their roots in the limits of formal systems. I will use interactions, self-organization, emergence, and balance to compare different aspects of complexity, ALife, and A.I. Even when the trajectory of the paper is influenced by my personal experience, the general questions posed (which outweigh the answers) hopefully will be useful to align efforts in these fields toward overcoming --- or accepting --- their limits.

This paper explores the increasing use of big data in education to create personalized learning experiences tailored to individual student needs. The trend, driven by vast amounts of educational data, requires the development of new techniques to optimize learning outcomes while addressing the ethical implications of data collection and utilization. The paper emphasizes the necessity for robust policy frameworks to ensure transparency and trust, based on clear guidelines and parental consent. It also discusses how educational institutions are updating curricula to prepare learners for a rapidly evolving workforce, incorporating contemporary AI technologies that impact education systems. The dual influence of AI presents both opportunities and challenges, necessitating a balanced and thoughtful approach. The paper advocates for a holistic strategy in integrating AI in education, emphasizing the importance of collaboration between the AI industry and other sectors to develop comprehensive policy frameworks for the ethical and effective use of AI in education.

In this paper, we investigate the proftability of investing in movies for retail investors, focusing on the return on investment (ROI) metric, which is calculated as revenue divided by budget. Our analysis encompasses 5000 movies from the movie database TMDB, examining factors such as production scale, genres, key actors, directors, and more. Additionally, we employ textual analysis techniques like Latent Semantic Analysis (LSA) on movie overviews and tags to incorporate movie themes. We then evaluate various supervised classifcation models including Logistic Regression, Random Forest, and Light Gradient Boosting (LightGBM), comparing their performance. Our fndings highlight the signifcance of production scale, team structure, and movie themes in identifying potential high-return opportunities for investors.

In this study, we explored the application of deep learning techniques for credit card fraud detection, aiming to improve the performance and reliability of anomaly detection methods in financial transactions. We first utilized the Isolation Forest algorithm, achieving a detection accuracy of 26% for the top 1000 transactions. Subsequently, we experimented with the Autoencoder algorithm, an unsupervised deep neural network model, which enhanced the detection accuracy to 33.6% in the best case despite some fluctuations. The results demonstrate deep learning models' strong feature extraction capability and adaptability, highlighting their potential to surpass traditional methods. However, the high imbalance in the dataset, with only 0.17% of transactions being fraudulent, poses a significant challenge. This study underscores the necessity for further experimentation and optimization of network structures and hyperparameters to achieve more stable and efficient fraud detection. The findings provide valuable insights and reference points for future research in financial fraud detection using deep learning methodologies.

In this work is proposed a new unsupervised method to evaluate the behavior of urban green areas in presence of heatwave scenarios analyzing the NDVI, NDMI and LST satellite data. To evaluate how these three indices characterize urban green areas during periods of heat waves, a reclassification of the data raster of each index is carried out. Rather than using standard classifications of the three indices, an unsupervised classification approach is proposed that uses the Elbow method to determine the optimal number of classes and the Jenks thematic classification method. Each class is assigned a Gaussian fuzzy set with means and standard deviations equal to that of the associated thematic class. The green urban areas are classified using zonal statistics operators and the degree of membership in the corresponding fuzzy set is used to assess the reliability of the classification. Finally, for each type of greenery the frequencies of belonging to the NDVI, NDMI and LST classes are analyzed to evaluate the behavior of the type of urban greenery during heatwaves. The framework was tested in an urban area consisting of the city of Naples (Italy). The results show that some types of greenery, such as deciduous forests and olive groves, are not very vulnerable to heatwave scenarios, unlike uncultivated areas which, instead, show critical behaviors during heatwaves.

In recent years, transformer-based models have gained prominence in multivariate long-term time series forecasting (LTSF), demonstrating significant advancements despite facing challenges such as high computational demands, difficulty in capturing temporal dynamics, and managing long-term dependencies. The emergence of LTSF-Linear, with its straightforward linear architecture, has notably outperformed transformer-based counterparts, prompting a reevaluation of the transformer's utility in time series forecasting. In response, this paper presents an adaptation of a recent architecture termed extended LSTM (xLSTM) for LTSF. xLSTM incorporates exponential gating and a revised memory structure with higher capacity that has good potential for LTSF. Our adopted architecture for LTSF termed as xLSTMTime surpasses current approaches. We compare xLSTMTime's performance against various state-of-the-art models across multiple real-world datasets, demonstrating superior forecasting capabilities. Our findings suggest that refined recurrent architectures can offer competitive alternatives to transformer-based models in LTSF tasks, potentially redefining the landscape of time series forecasting.

With the increasing incidents of fatal road injuries, there is an urgent need for developing effective road safety management systems. The study aims to develop predictive models based on machine learning to forecast the likelihood of road collisions depending on factors like weather, road condition, time, and driver behaviour in Chicago, USA. A machine learning approach has been applied to the crash dataset to evaluate the factors affecting the prevalence of road accidents. Python programming and the Jupyter Notebook platform have been used for performing descriptive statistics, correlation and three classification algorithms (Random Forest, K-Nearest Neighbor (KNN), Decision Tree and MLP Classification). Obtained accuracy of the KNN classifier is slightly higher than the other two classification models. The research explored insights into collision patterns related to roads, locations, and intersections. The study helps to increase road safety through targeted interventions with resource prioritisation, reducing the frequency and severity of traffic incidents by leveraging historical accident data with diverse spatial analysis techniques.

Malware detection is a crucial pillar of cybersecurity and aims to identify and neutralize malware threats. This study investigates how three deep learning algorithms - Convolutional Neural Net-works with One-Dimensional Architecture (CNN1D), Neural Networks (NN), and a hybrid model of one dimensional convolutional neural networks (CNN1D) and Long Short-Term Memory (LSTM) - help detect malware in Portable Executable working (.exe) files. Using a large dataset of tagged malware and benign samples, the effectiveness of these algorithms in distinguishing dangerous from harmless executables is extensively tested. In addition, the study evaluates the computational efficiency of the proposed methods for a variety of hardware configurations, including Central Processing Units (CPU), Graphics Processing Units (GPU), Raspberry Pi (IoT device), and Tensor Processing Units (TPU). Performance metrics provide insight into the scalability and effectiveness of the model across different hardware configurations. The results of this study have significant impact on the development of malware detection, particularly in the context of IoT security. By demonstrating the capabilities and limitations of CNN1D, Neural Networks, and CNN1D with LSTM models, the study provides practitioners with efficient and scalable solutions to real-world malware detection difficulties. In addition, insights from performance tests on various hardware platforms provide important assistance in developing robust security measures in IoT contexts, strengthening cybersecurity defenses in an ever-changing digital landscape.

As artificial intelligence (AI) makes significant headway in various arenas, the field of Speech-Language Pathology is at the precipice of experiencing a transformative shift towards automation. This study introduces QuickPic AAC, an AI-driven application designed to generate topic-specific displays from photographs in a "just-in-time" manner. Using QuickPic AAC, this study aimed to (a) determine which of two AI algorithms (NLG-AAC and GPT-3.5) results in greater specificity of vocabulary (i.e., percentage of vocabulary kept/deleted by clinician relative to vocabulary generated by QuickPic AAC; percentage of vocabulary modified); and to (b) evaluate perceived usability of QuickPic AAC among practicing speech-language pathologists. Results revealed that the GPT-3.5 algorithm consistently resulted in greater specificity of vocabulary and that speech-language pathologists expressed high user satisfaction for the QuickPic AAC application. These results support continued study of the implementation of QuickPic AAC in clinical practice and demonstrate the possibility of utilizing topic-specific displays as just-in-time supports.

Advanced Persistent Threats (APTs) are sophisticated cyberattacks aimed at stealing sensitive information or causing damage over an extended period. Detecting APTs is crucial for maintaining cybersecurity, and machine learning (ML) has emerged as a powerful tool in this domain. This paper explores the role of ML algorithms in detecting APTs, comparing their accuracy and effectiveness. We evaluate various algorithms, discuss the challenges in achieving superior accuracy, and suggest strategies for improvement. Our findings highlight the potential and limitations of ML in APT detection, emphasizing the need for continuous advancements in this field.

This study investigates the application of deep residual networks for predicting the dynamics of interacting three-dimensional rigid bodies. We present a framework combining a 3D physics simulator implemented in C++ with a deep learning model constructed using PyTorch. The simulator generates training data encompassing linear and angular motion, elastic collisions, fluid friction, gravitational effects, and damping. Our deep residual network, consisting of an input layer, multiple residual blocks, and an output layer, is designed to handle the complexities of 3D dynamics. We evaluate the network's performance using a dataset of 10,000 simulated scenarios, each involving 3-5 interacting rigid bodies. The model achieves a mean squared error of 0.015 for position predictions and 0.022 for orientation predictions, representing a 25% improvement over baseline methods. Our results demonstrate the network's ability to capture intricate physical interactions, with particular success in predicting elastic collisions and rotational dynamics. This work significantly contributes to physics-informed machine learning by showcasing the immense potential of deep residual networks in modeling complex 3D physical systems. We discuss our approach's limitations and propose future directions for improving generalization to more diverse object shapes and materials.

This paper presents an enhanced Fully Connected Neural Network (FCNN) framework that incorporates manifold learning and Renormalization Group (RG) methods. By integrating Christoffel symbols, the model adapts to the curvature and topology of the data manifold, improving learning from geometrically complex data. Additionally, RG methods enable effective multi-scale data processing. We demonstrate the enhanced model’s performance on synthetic and real-world datasets, showing improved accuracy compared to standard FCNNs, particularly in complex systems simulation.

This article comprehensively explores multiple aspects of cryptocurrencies and their price forecasting. Firstly, the article introduces the definition of cryptocurrency and its development process on a global scale, especially focusing on the launch of Facebook Libra and China's central bank digital currency, highlighting the importance and influence of digital currency in the global financial market. Subsequently, the article analyzes the advantages of digital currencies over traditional currencies, including improving economic transaction efficiency, reducing transaction costs and enhancing transaction transparency. Meanwhile, the article also explores the challenges and risks potentially brought by the development of digital currencies, such as regulatory uncertainty and market volatility. In this context, the article raises the importance of cryptocurrency price forecasting and introduces the forecasting models and techniques commonly used today. Finally, through specific experimental analysis, the effectiveness of using the deep learning model CNN-LSTM to predict the price of Bitcoin is demonstrated, and the directions of future research and optimization strategy are proposed. In summary, this paper comprehensively presents the research status and prospects of cryptocurrency and its price prediction field through systematic introduction and analysis.

In today's highly digitized and globalized financial environment, the need to protect payment systems from risk is more urgent than ever. Artificial intelligence (AI) technology is rapidly becoming a key tool in this space, with machine learning algorithms, big data analytics, and real-time monitoring enabling AI to effectively detect and prevent fraudulent activity, optimize risk management processes, and deliver intelligent services. AI not only improves the security and efficiency of payment systems, but also significantly improves customer satisfaction and loyalty. With the continuous development of AI technology and financial payment technology, the financial payment industry will usher in more innovations and changes in the future, further optimizing the payment process, improving payment efficiency, and ensuring payment security.

This article presents a succinct overview of the progress, obstacles, and uses of semantic segmentation and instance segmentation within the forestry domain. The objective of this review is to conduct a critical analysis of the current literature pertaining to segmentation techniques and provide a methodical summary of their impact on forestry-related activities, including but not limited to tree species classification, forest inventory, and ecological monitoring such as retrieval of dominant tree species. Through the process of synthesizing pivotal discoveries from multiple studies, this comprehensive analysis provides valuable perspectives on the present status of research and highlights prospective areas for further exploration. The primary topics addressed encompass the approach employed for executing the examination, the fundamental discoveries associated with semantic segmentation and instance segmentation in the domain of forestry, and the ramifications of these discoveries for the discipline. The results indicate that the utilization of semantic segmentation methods has been efficacious in the field of forestry for the precise identification of tree species. Such methods also aid in tracking of deforested regions over the course of time by separating other land-use classes from forested regions. Additionally, the employment of instance segmentation techniques exhibits potential in the demarcation of individual trees. Instance segmentation offers promising results due to deep learning models based on forest point clouds. However, several challenges persist in the successful implementation of semantic segmentation methods such as the presence of occlusions, overlapping branches, and intricate structures hampers the accurate segmentation of trees. Additionally, instance segmentation approaches that utilize models are mostly trained by using laser scanning data based on forest types which are typically trained on specific laser scanning data and forest types that create limitations in generalization from high to low resolution point clouds. Due to this reason, the existing approaches often struggle with handling these complex structures, leading to the need for manual methods for extracting measurements from forest point clouds. The review culminates by underscoring the necessity for additional research to tackle current obstacles and augment the precision and relevance of segmentation methodologies in the field of forestry. In general, the present article provides a significant reference for scholars and professionals who are interested in utilizing segmentation techniques in the field of forestry.

Astronomy is entering an unprecedented era of Big Data science, driven by missions like the ESA's Gaia telescope, which aims to map the Milky Way in three dimensions. Gaia's vast dataset presents a monumental challenge for traditional analysis methods. The sheer scale of this data exceeds the capabilities of manual exploration, necessitating the utilization of advanced computational techniques. In response to this challenge, we developed a novel approach leveraging deep learning to estimate the metallicity of fundamental mode (ab-type) RR Lyrae stars from their light curves in the Gaia optical G-band. Our study explores applying deep learning techniques, particularly advanced neural network architectures, in predicting photometric metallicity from time-series data. Moreover, our deep learning models demonstrated notable predictive performance over experimentation and analysis. Through GRU-based models we achieved a low mean absolute error of 0.056 and a high $R^2$ regression performance of 0.9401 measured by cross-validation, showcasing the effectiveness of our approach in accurately estimating metallicity values. Our work underscores the importance of deep learning in astronomical research, particularly with large datasets from missions like Gaia. By harnessing the power of deep learning methods, we can provide precision in analyzing vast datasets, contributing to more precise and comprehensive insights into complex astronomical phenomena.

Cryptocurrencies have now become an emerging blockchain-based payment technology. Users’ identities on such networks are pseudo-anonymous in that all transactions made from an address are transparent and searchable by anyone; to preserve their privacy, users often use many different addresses. In recent years, some studies have been conducted regarding analyzing clusters of Bitcoin addresses that, according to certain heuristics, belong to the same entity. Such action allows law enforcement to have relevant information in cases where cryptocurrencies are used for illegal trafficking. Clustering methods based on a single heuristic do not allow the clustering of many addresses in a complete and accurate manner. This paper proposes Bitcoin Address Clustering based on multiple Heuristics (BACH), a tool that uses three different clustering heuristics to identify clusters of Bitcoin addresses, which are displayed through a three-dimensional graph. Based on the results, several analyses were conducted, including a comparison with Wallet Explorer, an address clustering tool similar to BACH. BACH, in addition to introducing the innovative feature of graphical visualization of the internal structure of clusters, is shown to improve address aggregation through the joint use of three heuristics for clustering.

This study investigates the latest advancements in utilizing machine learning techniques for predicting financial market movements. Financial markets are complex systems influenced by numerous factors, making accurate predictions challenging but highly valuable. The research explores various machine learning approaches, including supervised learning, unsupervised learning, reinforcement learning, and deep learning, to enhance the accuracy and reliability of financial market forecasts. The study provides a comprehensive review of current methodologies, highlighting key innovations and their applications in financial market prediction. Techniques such as neural networks, support vector machines, random forests, and ensemble methods are examined in the context of their ability to analyze vast amounts of financial data, identify patterns, and generate actionable insights. Additionally, the research delves into the integration of alternative data sources, such as social media sentiment, news articles, and macroeconomic indicators, to improve prediction models. Through empirical analysis and case studies, the study demonstrates the effectiveness of advanced machine learning models in predicting stock prices, currency exchange rates, and commodity prices. It also discusses the challenges associated with these techniques, including overfitting, data quality issues, and the need for interpretability in model outputs. Findings suggest that machine learning techniques can significantly enhance the predictive power of financial models, offering more accurate and timely insights for investors and financial analysts. The study concludes with recommendations for future research directions, emphasizing the importance of model transparency, the incorporation of diverse data sets, and the continuous adaptation of machine learning algorithms to evolving market conditions.

This study presents a strategic approach to developing applications focused on implementing Large Language Models using the Langchain framework in Python. Three language models are highlighted: GPT-3.5 (turbo mode), LLaMA 2, and Vicuna 7B, each with their distinctive features and capabilities. The methodology used is described in detail, including data extraction from medical reports using zero-shot prompting data extraction techniques, interaction with language models, and structured storage of results. The performance of the models in data extraction is evaluated, presenting metrics such as precision, recall, and F1 score. The results demonstrate high model capability in extracting information, although areas for improvement are identified, particularly in data extraction precision. In conclusion, the efficacy of the models in extracting information from medical histories is not considerably acknowledged, with an emphasis on the importance of improving precision and increasing the volume of trained data for future research in healthcare digitalization.

In This paper we introduce the integration of Renormalization Group (RG) methods with Deep Q-Networks (DQNs) to improve reinforcement learning in high-dimensional state spaces. RG methods provide multi-scale analysis, enhancing state representation, learning stability, and exploration. The proposed RG-DQN algorithm uses hierarchical Q-value estimation and multi-scale representations, demonstrating superior performance on synthetic genomic data compared to traditional DQNs.}

Multimodal emotion classification (MEC) involves analyzing and identifying human emotions by integrating data from multiple sources, such as audio, video, and text. This approach leverages the complementary strengths of each modality to enhance the accuracy and robustness of emotion recognition systems. However, one significant challenge is effectively integrating these diverse data sources, each with unique characteristics and levels of noise. Additionally, the scarcity of large, annotated multimodal datasets in Bangla limits the training and evaluation of models. In this work, we unveiled a pioneering multimodal Bangla dataset, MAViT-Bangla (Multimodal Audio Video Text Bangla dataset). This dataset, comprising 1002 samples across audio, video, and text modalities, is a unique resource for emotion recognition studies in the Bangla language. It features emotional categories such as anger, fear, joy, and sadness, providing a comprehensive platform for research. Additionally, we developed a framework for emotion recognition that employs a cross-modal attention mechanism among unimodal features. This mechanism fosters the interaction and fusion of features from different modalities, enhancing the model’s ability to capture nuanced emotional cues. The effectiveness of this approach was demonstrated by achieving an F1-score of 0.64, a significant improvement over unimodal methods.

Physics-informed neural networks can be trained to serve as surrogate models by learning how spatial physical fields develop with change of parameters. This is an advantage over traditional numerical methods that can only be used to calculate a single solution field for a single discrete setting of a parameter. While many studies report accurate results of physics-informed deep learning of laminar flows, evaluations of surrogate modeling of high Reynolds number flows remain sparse. To contribute to this question, we here explore the capabilities and limits of physics-informed deep learning to solve the Reynolds-averaged Navier-Stokes equations for the flow around two airfoils under variable angles of attack. No labeled training data were provided and a single network was trained to simultaneously learn several solution fields in a single training run. The network captured essential flow features such as boundary layers and high and low pressure regions. Moreover, the qualitative correlations of lift and drag with the angle of attack and the higher drag of the thicker airfoil were captured. However, comparisons with reference simulations revealed an underestimation of the shear layer gradients as well as deviations of the fields' extrema. Thus, we also debate potential future improvements to increase the accuracy of the predicted flow. This work offers a prospect on the capabilities and current limitations of physics-informed neural networks for surrogate modeling of high Reynolds number flows. We argue that the method is promising and this work is aimed to encourage further research and development of the method when applied to Reynolds-averaged flows at high Reynolds numbers.

This study investigates the impact of AI regulations and adoption on labor markets and employment in the USA. In order to achieve the study's objective, a robust econometric approach was employed using Principal Components Analysis (PCA) to address multicollinearity among the variables. The regression analysis incorporated principal components representing overall AI adoption and education (PC1), AI innovation and academic output (PC2), economic growth (PC3), and unemployment and AI strategies (PC4). The analysis utilized heteroskedasticity-robust standard errors to ensure reliable coefficient estimates and tested for stationarity to confirm the stability of the time series data. Data were collected from 2010 to 2022, encompassing key AI-related and economic indicators. The results reveal that higher AI adoption and education levels initially lead to job displacement, negatively impacting labor market metrics. Similarly, AI innovation and economic growth driven by AI do not immediately translate into job creation, reflecting transitional challenges. However, the strategic implementation of AI significantly mitigates these adverse effects and enhances employment conditions. The study finds that comprehensive AI strategies, robust regulatory frameworks, and effective workforce retraining and upskilling programs are essential for supporting labor market stability and promoting employment growth. Based on the findings, the study recommended that there is the need to develop and implement comprehensive AI strategies that include robust regulatory frameworks to support workforce transitions. Policies should focus on retraining and upskilling programs to help displaced workers adapt to new AI-driven roles.

This paper explores the application of machine learning in financial time series analysis, focusing on predicting trends in financial enterprise stocks and economic data. It begins by distinguishing stocks from stocks and elucidates risk management strategies in the stock market. Traditional statistical methods such as ARIMA and exponential smoothing are discussed in terms of their advantages and limitations in economic forecasting. Subsequently, the effectiveness of machine learning techniques, particularly LSTM and CNN-BiLSTM hybrid models, in financial market prediction is detailed, highlighting their capability to capture nonlinear patterns in dynamic markets. The study demonstrates the advancements in predictive accuracy and robustness achieved by deep learning methods through empirical analysis and model validation. The findings contribute significantly to academic discourse and offer practical insights for investors, financial analysts, and policymakers navigating market volatility and optimizing investment strategies. Finally, the paper outlines prospects for machine learning in financial forecasting, laying a theoretical foundation and methodological framework for achieving more precise and reliable economic predictions.

This survey provides a comprehensive overview of in-context learning (ICL) in large language models (LLMs), a phenomenon where models can adapt to new tasks without parameter updates by leveraging task-relevant information within the input context. We explore the definition and mechanisms of ICL, investigate the factors contributing to its emergence, and discuss strategies for optimizing and effectively utilizing ICL in various applications. Through a systematic review of recent literature, we first clarify what ICL is, distinguishing it from traditional fine-tuning approaches and highlighting its unique characteristics. We then delve into the underlying causes of ICL, examining theories ranging from implicit meta-learning during pre-training to the emergence of task vectors in LLMs. The survey also covers various approaches to enhance ICL performance, including prompt engineering techniques, demonstration selection strategies, and methods for improving generalization across diverse tasks. Additionally, we discuss the limitations and challenges of ICL, such as its sensitivity to demonstration ordering and potential biases. By synthesizing findings from numerous studies, we aim to provide researchers and practitioners with a clear understanding of the current state of ICL research, its practical implications, and promising directions for future investigation. This survey serves as a valuable resource for those seeking to leverage ICL capabilities in LLMs and contributes to the ongoing discourse on the remarkable adaptability of these models.

This research addresses the challenge of integrating fairness considerations into optimization problems, focusing on the development of multi-objective optimization techniques that balance traditional efficiency metrics with fairness constraints. In many algorithmic decision-making processes, optimizing for efficiency alone can inadvertently perpetuate or exacerbate biases, leading to inequitable outcomes. This study explores how data analysis and machine learning can be leveraged to identify and mitigate these biases, ensuring that optimization models yield both efficient and fair solutions. The research presents a framework for incorporating fairness constraints into multi-objective optimization, utilizing various fairness metrics such as demographic parity, equal opportunity, and disparate impact. It investigates the application of this framework across different domains, including resource allocation in healthcare systems, loan approvals in finance, and personalized learning platforms in education. By analyzing real-world data sets, the study demonstrates how fairness-aware optimization can lead to more equitable outcomes without significantly compromising efficiency. Empirical results from case studies show that multi-objective optimization with fairness constraints can effectively balance the trade-off between efficiency and equity. The research also discusses the computational challenges and ethical considerations associated with implementing fairness constraints in optimization models. Strategies for addressing these challenges, such as regularization techniques and fairness-aware machine learning algorithms, are evaluated and presented. Findings highlight the importance of incorporating fairness into optimization processes to prevent algorithmic biases and promote social equity. The study concludes with practical recommendations for policymakers, practitioners, and researchers on adopting fairness-aware optimization techniques to ensure that algorithmic decision-making processes are both efficient and just.