Crime Prediction based on Classification Approaches

This study assesses and contrasts the effectiveness of four supervised learning algorithms which are Decision Tree, Extra Tree, Random Forest, and XGBoost in predicting crime types by utilizing the San Francisco crime dataset. The main goal is to identify the most efficient model for forecasting crime incidents based on past data. The results demonstrate that the Random Forest algorithm surpasses the other algorithms with an accuracy rate of 34%. This research highlights the importance of machine learning classifiers in helping security agencies optimize resources and improve public safety by accurately predicting crimes. The findings suggest that machine learning methods can greatly enhance crime prediction abilities and offer valuable insights for proactive policing strategies.