You are currently viewing a beta version of our website. If you spot anything unusual, kindly let us know.
Altmetrics
Downloads
89
Views
120
Comments
0
This version is not peer-reviewed
Study | Focus | Data Used | Models Evaluated | Key Findings |
---|---|---|---|---|
[5] | Crime category prediction | Crime and Communities dataset, UCI | RF, SVM | RF outperforms SVM with 99% accuracy. |
[6] | Crime category prediction in Chicago | Chicago crime data | Decision Trees, Naive Bayes | Decision Trees outperform Naive Bayes with 91.68% accuracy. |
[7] | Crime prediction in India | NCRB Indian criminal records | SVM, J48, SMO, Naïve Bayes, Bagging, Random Forest | Ensemble method (SBCPM) shows superior performance with 99.5% accuracy. |
[8] | Crime prediction in Dubai | Dubai crime data | Random Forest, KNN, SVM, ANN, Naïve Bayes, Decision Tree | KNN outperforms others with 78.47% accuracy. |
[9] | Crime prediction in urban areas | Chicago Police Department’s CLEAR system data | Random Forest, KNN, AdaBoost, Neural Networks | Neural Network shows highest accuracy at 90.77%. |
[10] | Automated crime report analysis | San Francisco Crime dataset | Random Forest | Achieved 86.5% accuracy, precision, recall, F-score, AUC of 0.98. |
[11] | Criminal activity prediction in San Francisco | San Francisco criminal incidents data | Decision Tree, KNN, Random Forest, AdaBoost | Random Forest with random undersampling achieved 99.16% accuracy. |
[12] | Crime trend analysis in Bangladesh | Bangladesh crime data | KNN, Naive Bayes, Linear Regression | KNN outperforms with 76.92% accuracy in predicting safe routes. |
[13] | Crime data prediction | Chicago and Los Angeles crime data | LSTM, ARIMA, Logistic Regression, SVM, Naive Bayes, KNN, Decision Tree, MLP, Random Forest, XGBoost | LSTM effective in time series forecasting; ARIMA predicts crime trends. |
[14] | Crime prediction with machine learning | San Francisco crime data | Naive Bayes, Random Forest, Gradient Boosting Decision Trees | Gradient Boosting Decision Trees show superior performance with 98.5% accuracy. |
Current Work | Crime category prediction in Montreal | Montreal crime data (2015-2023) | XGBoost, Decision Trees, Random Forest | XGBoost outperforms others with superior precision, accuracy, recall, and F1 score. Deployed as a web application. |
Attribute | Description |
---|---|
Crime category | The nature of the event:
|
Date | Date of event report to MCPS in YYYY-MM-DD format. |
Time of Day | Time of day the event was reported to MCPS:
|
Police District Number | Police district number for the neighborhood where the incident occurred. |
X | Geospatial position according to the MTM8 projection (SRID 2950). |
Y | Geospatial position according to the MTM8 projection (SRID 2950). |
Latitude | Geographical position of the event at an intersection according to the WGS84 geodetic reference system. |
Longitude | Geographical position of the event at an intersection according to the WGS84 geodetic reference system. |
Attribute | Number of Missing Data | Percentage Missing (%) |
---|---|---|
X | 47193 | 16.854 |
Y | 47193 | 16.854 |
Longitude | 47193 | 16.854 |
Latitude | 47193 | 16.854 |
Police District Number | 5 | 0.0018 |
Crime Category | Numeric Encoding |
---|---|
Theft from/in Motor Vehicle | 0 |
Break and Enter | 1 |
Mischief | 2 |
Motor Vehicle Theft | 3 |
Robbery | 4 |
Offenses Causing Death | 5 |
Time | Numeric Encoding |
---|---|
Day | 0 |
Evening | 1 |
Night | 2 |
Balancing Algorithm | Accuracy |
---|---|
SMOTE | 0.588510 |
SMOTE-Tomek | 0.474414 |
SMOTE-ENN | 0.632867 |
ADASYN | 0.902227 |
Class | Precision | Recall | F1-score | Support | Accuracy |
---|---|---|---|---|---|
Results for XGBoost | |||||
Theft from/in Motor Vehicle | 0.82 | 0.62 | 0.71 | 1026 | |
Break and Enter | 0.81 | 0.72 | 0.76 | 1848 | |
Mischief | 0.81 | 0.65 | 0.72 | 1807 | |
Motor Vehicle Theft | 0.84 | 0.79 | 0.81 | 2601 | |
Robbery | 0.88 | 0.96 | 0.91 | 8894 | |
Offenses Causing Death | 0.99 | 1.00 | 0.99 | 13669 | |
Weighted Avg | 0.91 | 0.92 | 0.91 | 29845 | 0.92 |
Results for Decision Tree | |||||
Theft from/in Motor Vehicle | 0.53 | 0.48 | 0.51 | 1026 | |
Break and Enter | 0.61 | 0.55 | 0.58 | 1848 | |
Mischief | 0.57 | 0.55 | 0.56 | 1807 | |
Motor Vehicle Theft | 0.70 | 0.68 | 0.69 | 2601 | |
Robbery | 0.84 | 0.86 | 0.85 | 8894 | |
Offenses Causing Death | 0.98 | 0.99 | 0.99 | 13669 | |
Weighted Avg | 0.85 | 0.86 | 0.85 | 29845 | 0.86 |
Results for RF | |||||
Theft from/in Motor Vehicle | 0.82 | 0.30 | 0.44 | 1026 | |
Break and Enter | 0.79 | 0.32 | 0.46 | 1848 | |
Mischief | 0.88 | 0.37 | 0.52 | 1807 | |
Motor Vehicle Theft | 0.81 | 0.63 | 0.71 | 2601 | |
Robbery | 0.72 | 0.92 | 0.81 | 8894 | |
Offenses Causing Death | 0.93 | 1.00 | 0.96 | 13669 | |
Weighted Avg | 0.84 | 0.84 | 0.82 | 29845 | 0.84 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
supplementary.csv (24.12MB )
Submitted:
17 June 2024
Posted:
19 June 2024
You are already at the latest version
supplementary.csv (24.12MB )
This version is not peer-reviewed
Submitted:
17 June 2024
Posted:
19 June 2024
You are already at the latest version
Study | Focus | Data Used | Models Evaluated | Key Findings |
---|---|---|---|---|
[5] | Crime category prediction | Crime and Communities dataset, UCI | RF, SVM | RF outperforms SVM with 99% accuracy. |
[6] | Crime category prediction in Chicago | Chicago crime data | Decision Trees, Naive Bayes | Decision Trees outperform Naive Bayes with 91.68% accuracy. |
[7] | Crime prediction in India | NCRB Indian criminal records | SVM, J48, SMO, Naïve Bayes, Bagging, Random Forest | Ensemble method (SBCPM) shows superior performance with 99.5% accuracy. |
[8] | Crime prediction in Dubai | Dubai crime data | Random Forest, KNN, SVM, ANN, Naïve Bayes, Decision Tree | KNN outperforms others with 78.47% accuracy. |
[9] | Crime prediction in urban areas | Chicago Police Department’s CLEAR system data | Random Forest, KNN, AdaBoost, Neural Networks | Neural Network shows highest accuracy at 90.77%. |
[10] | Automated crime report analysis | San Francisco Crime dataset | Random Forest | Achieved 86.5% accuracy, precision, recall, F-score, AUC of 0.98. |
[11] | Criminal activity prediction in San Francisco | San Francisco criminal incidents data | Decision Tree, KNN, Random Forest, AdaBoost | Random Forest with random undersampling achieved 99.16% accuracy. |
[12] | Crime trend analysis in Bangladesh | Bangladesh crime data | KNN, Naive Bayes, Linear Regression | KNN outperforms with 76.92% accuracy in predicting safe routes. |
[13] | Crime data prediction | Chicago and Los Angeles crime data | LSTM, ARIMA, Logistic Regression, SVM, Naive Bayes, KNN, Decision Tree, MLP, Random Forest, XGBoost | LSTM effective in time series forecasting; ARIMA predicts crime trends. |
[14] | Crime prediction with machine learning | San Francisco crime data | Naive Bayes, Random Forest, Gradient Boosting Decision Trees | Gradient Boosting Decision Trees show superior performance with 98.5% accuracy. |
Current Work | Crime category prediction in Montreal | Montreal crime data (2015-2023) | XGBoost, Decision Trees, Random Forest | XGBoost outperforms others with superior precision, accuracy, recall, and F1 score. Deployed as a web application. |
Attribute | Description |
---|---|
Crime category | The nature of the event:
|
Date | Date of event report to MCPS in YYYY-MM-DD format. |
Time of Day | Time of day the event was reported to MCPS:
|
Police District Number | Police district number for the neighborhood where the incident occurred. |
X | Geospatial position according to the MTM8 projection (SRID 2950). |
Y | Geospatial position according to the MTM8 projection (SRID 2950). |
Latitude | Geographical position of the event at an intersection according to the WGS84 geodetic reference system. |
Longitude | Geographical position of the event at an intersection according to the WGS84 geodetic reference system. |
Attribute | Number of Missing Data | Percentage Missing (%) |
---|---|---|
X | 47193 | 16.854 |
Y | 47193 | 16.854 |
Longitude | 47193 | 16.854 |
Latitude | 47193 | 16.854 |
Police District Number | 5 | 0.0018 |
Crime Category | Numeric Encoding |
---|---|
Theft from/in Motor Vehicle | 0 |
Break and Enter | 1 |
Mischief | 2 |
Motor Vehicle Theft | 3 |
Robbery | 4 |
Offenses Causing Death | 5 |
Time | Numeric Encoding |
---|---|
Day | 0 |
Evening | 1 |
Night | 2 |
Balancing Algorithm | Accuracy |
---|---|
SMOTE | 0.588510 |
SMOTE-Tomek | 0.474414 |
SMOTE-ENN | 0.632867 |
ADASYN | 0.902227 |
Class | Precision | Recall | F1-score | Support | Accuracy |
---|---|---|---|---|---|
Results for XGBoost | |||||
Theft from/in Motor Vehicle | 0.82 | 0.62 | 0.71 | 1026 | |
Break and Enter | 0.81 | 0.72 | 0.76 | 1848 | |
Mischief | 0.81 | 0.65 | 0.72 | 1807 | |
Motor Vehicle Theft | 0.84 | 0.79 | 0.81 | 2601 | |
Robbery | 0.88 | 0.96 | 0.91 | 8894 | |
Offenses Causing Death | 0.99 | 1.00 | 0.99 | 13669 | |
Weighted Avg | 0.91 | 0.92 | 0.91 | 29845 | 0.92 |
Results for Decision Tree | |||||
Theft from/in Motor Vehicle | 0.53 | 0.48 | 0.51 | 1026 | |
Break and Enter | 0.61 | 0.55 | 0.58 | 1848 | |
Mischief | 0.57 | 0.55 | 0.56 | 1807 | |
Motor Vehicle Theft | 0.70 | 0.68 | 0.69 | 2601 | |
Robbery | 0.84 | 0.86 | 0.85 | 8894 | |
Offenses Causing Death | 0.98 | 0.99 | 0.99 | 13669 | |
Weighted Avg | 0.85 | 0.86 | 0.85 | 29845 | 0.86 |
Results for RF | |||||
Theft from/in Motor Vehicle | 0.82 | 0.30 | 0.44 | 1026 | |
Break and Enter | 0.79 | 0.32 | 0.46 | 1848 | |
Mischief | 0.88 | 0.37 | 0.52 | 1807 | |
Motor Vehicle Theft | 0.81 | 0.63 | 0.71 | 2601 | |
Robbery | 0.72 | 0.92 | 0.81 | 8894 | |
Offenses Causing Death | 0.93 | 1.00 | 0.96 | 13669 | |
Weighted Avg | 0.84 | 0.84 | 0.82 | 29845 | 0.84 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
Yasmine Lamari
et al.
IJGI,
2020
Irina Matijosaitiene
et al.
IJGI,
2018
Dongyoung Kim
et al.
Sustainability,
2021
© 2024 MDPI (Basel, Switzerland) unless otherwise stated