Abstract: The automotive sector is changing fast with more integration of advanced communication technologies and further connectivity. The modern vehicle is already a collection of diverse Electronic Control Units (ECU) communicating over interconnected networks that decide critical functionalities such as engine control, braking, and entertainment. However, this increasing complexity also introduces major cybersecurity risks, including network vulnerabilities like IP spoofing, message replay, and denial-of-service(DoS) attacks, besides software vulnerabilities due to coding errors in unsafe languages like C/C++. These are serious threats to vehicle operational reliability, passenger safety, and data integrity, making robust automotive security a critical concern. This paper explores the application of CHERI(Capability Hardware Enhanced RISC Instructions) in enhancing the security of Intrusion Detection Systems(IDS) in automotive networks. CHERI introduces fine-grained memory protection mechanisms that mitigate software vulnerabilities by enforcing spatial memory safety and preventing unauthorized access to critical data. Moreover, CHERI secures IDS rule configurations from network-based threats, such as manipulation of rules and spoofing attacks, by utilizing strict memory bounds and capability-based access controls. This work experimentally demonstrates that CHERI-enhanced IDSs are highly effective in identifying and mitigating spoofing and IDS rule manipulation attacks, ensuring the integrity of rules even against attackers using forged traffic with legitimate-looking source IP addresses. The results highlight CHERI’s hardware-enforced security model as a robust solution for preventing network and software-level exploits without compromising performance while maintaining compatibility with automotive-friendly programming languages like C/C++. This study underscores the critical importance of integrating CHERI and other hardware-based security frameworks into connected and autonomous vehicles to address emerging cybersecurity challenges and build a safer automotive ecosystem.

Abstract: This paper extends the well-known transfinite interpolation formula, which was developed in late sixties by the applied mathematician William Gordon at the premises of General Motors as an extension of the pre-existed Coons interpolation formula. Here, a conjecture is formulated, which claims that the meaning of the involved blending functions can be enlarged, so that to include any linear independent and complete set of functions, including piecewise-linear, trigonometric functions, Bernstein polynomials, B-splines, NURBS and so on. In this sense, NURBS-based isogeometric analysis and aspects of T-splines may be considered as special cases. Applications are provided for the accuracy in the interpolation through the L2-error norm, of closed-formed functions prescribed at the nodal points of the transfinite patch, which represent the solution of partial differential equations under boundary conditions of Dirichlet type.

Abstract: This research investigates the automatic detection of humour. The basic structure of humor consists in the punchline and contextual meaning leading to the joke. Although a joke's pith lies in the punch line, it would be difficult to interpret one that doesn't have enough context. One of the biggest obstacles to humour recognition is that it is subjective. As a result, difficulties arise because it can be difficult to determine whether something should be considered humourous or not. Nonetheless, humour undoubtedly becomes easier to understand when this aspect of everyday communication is incorporated into machines. To investigate the automatic detection of humour, this research was carried out on 200k samples and uses deep learning architecture; RNN(LSTM) in conjunction with two pre-trained embeddings; GLOVE and FASTTEXT, BERT, DistilBERT and a machine learning classifier Naïve Bayes to classify and predict humour. The experimental findings suggest all models are effective for humour detection in text, with BERT achieving the highest accuracy of 97% as a result, explains the importance of contextual approach in detecting humour.

Abstract: This study presents an advanced pothole detection system utilizing ensemble learning (YOLOv9 instance segmentation and Mask R-CNN) and a Multi-Criteria Decision Making (MCDM) framework to improve detection reliability. The system combines YOLOv9 for rapid instance segmentation and Mask R-CNN for precise segmentation, experimenting with adjusted confidence thresholds to enhance detection rates in challenging scenarios. For Yolov9 instance segmentation model achieved a mean Average Precision (mAP) of 0.908 at 0.5 IoU and an F1-score of 0.58 at a confidence threshold of 0.282. The F1-confidence curve highlights a strong balance between precision and recall, but further work is needed to ensure generalization. Dynamic weights are used to merge outputs, leveraging the strengths of both models. The MCDM framework refines detections by evaluating pothole features such as size, position, and shape. While the system demonstrates high detection accuracy of 20%, narrowly and over-specific defined MCDM criteria may lead to overfitting, limiting adaptability to diverse conditions. The study underscores the importance of balancing accuracy and adaptability for reliable performance in varied environments.

Abstract: This study implements the proposed integration of machine learning (ML) techniques into the ISO/IEC 27043:2015 international standard processes using a hypothetical case scenario for a smart building. ISO/IEC 27043:2015 does not currently incorporate ML techniques. Incorporating these techniques into ISO/IEC 27043:2015 can improve the efficiency of the processes and reduce time and human effort by automating some manual tasks of the readiness processes. This research presents a case study for the smart building dataset, applying ML techniques to implement the ML readiness model in the ISO/IEC 27043:2015 standard. It compares the results of implementing ML techniques. These results indicate how the smart environment data can be proactively analysed and classified. These techniques will enable investigators to access the information to investigate such environments.

Abstract:

The Hartman-Grobman Theorem plays a pivotal role in the qualitative analysis of dynamical systems, providing insights into the behavior of systems near hyperbolic equilibrium points through linear approximations. This paper presents an in-depth exploration of the theorem, clarifying its technical stipulations and demonstrating its application with practical examples. We begin by defining key concepts integral to dynamical systems such as equilibrium points, linearization, and the Jacobian matrix. Subsequent sections discuss the conditions under which the theorem applies, particularly focusing on hyperbolicity and the importance of eigenvalues in determining system stability. Additionally, the notion of topological conjugacy is examined to illustrate how nonlinear and linear system trajectories correlate qualitatively. We further investigate the concept of Lipschitz continuity and its relevance to the theorem's applicability. Through illustrative examples, including simple linear systems and more complex saddle points, we underscore the theorem's utility in simplifying the understanding of nonlinear dynamics. This comprehensive coverage of the theme not only elucidates the fundamental aspects of the Hartman-Grobman Theorem but also highlights its significant implications for predicting and analyzing system behavior in various scientific and engineering applications.

Abstract: The P versus NP problem is a cornerstone of theoretical computer science, asking whether problems that are easy to check are also easy to solve. "Easy" here means solvable in polynomial time, where the computation time grows proportionally to the input size. While this problem's origins can be traced to John Nash's 1955 letter, its formalization is credited to Stephen Cook and Leonid Levin. Despite decades of research, a definitive answer remains elusive. Central to this question is the concept of NP-completeness. If even one NP-complete problem, like SAT, could be solved efficiently, it would imply that all NP problems could be solved efficiently, proving P=NP. This research proposes a groundbreaking claim: SAT, traditionally considered NP-complete, can be solved in polynomial time, establishing the equivalence of P and NP.

Abstract:

Urban cycling is a sustainable mode of transportation in large cities, and it offers many advantages. It is an eco-friendly means of transport that is accessible to the population and easy to use. Additionally, it is more economical than other means of transportation. Urban cycling is beneficial for physical health and mental well-being. Achieving sustainable mobility and the evolution towards smart cities demands a comprehensive analysis of all the essential aspects that enable their inclusion. Road safety is particularly important, which must be prioritized to ensure safe transportation and reduce the incidence of road accidents. In order to help reduce the number of accidents that urban cyclists are involved in, this work proposes an alternative solution in the form of an intelligent computational assistant that utilizes Simplified Machine Learning (SML) to detect potential risks of unexpected collisions. This technological approach serves as a helpful alternative to the current problem. Through our methodology, we were able to identify the problem involved in the research, design and development of the solution proposal, collect and analyze data, and obtain preliminary results. These results experimentally demonstrate how the proposed model outperforms most state-of-the-art models that use a metric learning layer for small image sets.

Abstract: In experimental research, animal welfare should always be of the highest priority. Currently, physical in-person observations are the standard. This is time consuming, and results are subjective. Video-based machine learning models to monitor experimental pigs provides a continuous and objective observation method for animal misthrive detection. The aim of this study was to develop and validate a pig tracking technology, using video-based data in a machine learning model to analyze posture and activity level of experimental pigs living in single-pig pens. A research prototype was created using a microcomputer and a ceiling mounted camera for live recording based on the obtained images from the experimental facility and a combined model was created based on the Ultralytics YOLOv8n for object detection trained on the obtained images. As a second step, the Lucas-Kanade sparse optical flow technique for movement detection was applied. The resulting model successfully classified whether individual pigs were laying, standing, or walking. The validation test showed an accuracy of 90.66%, precision of 90.91%, recall of 90.66%, and a correlation coefficient of 84.53% compared with observed ground truth. In conclusion, the model demonstrates how machine learning can be used to monitor experimental animals potentially to improve animal welfare.

Abstract: Human Activity Recognition (HAR) and gender recognition have become pivotal in advancing intelligent systems, as they enable tailored user experiences and enhance automation in healthcare, security, and personalized technology. By accurately identifying human activities and gender, systems can proactively adapt to user needs, improving human-computer interaction, reducing response times in emergency detection, and enhancing the quality of life in smart home and assistive living applications. In this study, a Kolmogorov-Arnold Network model was created to predict daily living activities and the gender of the person performing an activity. The proposed Kolmogorov-Arnold Network (KAN) classifier outperformed the previous studies with 94.5% and 95.6% in terms of overall accuracy for multi-class HAR and gender recognition tasks, respectively. Additionally, the Chi-square test results demonstrated that there is a statistically significant difference between the performances. It can be concluded that KAN method is a robust classifier especially for detecting activities that have a minor number of samples on the utilized dataset.