Prediction of Theft Risk Areas using Machine Learning Algorithms

This study explores the prediction of robbery risk areas in the Callao region by utilizing machine learning algorithms. It employs geospatial and temporal data, including information about location, time, and security modalities. The algorithms used in this research include K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Random Forest, Logistic Regression, Naive Bayes, Gradient Boosting, and XGBoost. The methodology encompassed data collection, preprocessing, model construction, and evaluation. The results indicate that XGBoost is the most effective model for classifying active and inactive robbery incidents, achieving an impressive accuracy of 90.10%. This model outperformed other methods in terms of predictive accuracy. This machine learning-based approach shows promise as a valuable tool for preventive urban security analysis, laying the groundwork for more proactive security policies in high-crime areas.