Optimization Analysis of Crime Rate Prediction using K-mean and Machine Learning Algorithm

Abstract

Crime rate prediction plays a crucial role in enhancing public safety and supporting proactive law enforcement strategies. With the rapid growth of urbanization and the availability of large-scale crime datasets, traditional statistical approaches are often inadequate for capturing complex spatial and temporal crime patterns. This study presents an optimization analysis of crime rate prediction using K-means clustering and machine learning algorithms. The proposed framework integrates unsupervised clustering with supervised learning models to improve prediction accuracy and reduce data complexity. K-means clustering is employed as a preprocessing step to group crime data into homogeneous clusters based on spatial and behavioral similarities, enabling effective identification of crime hotspots. Subsequently, machine learning algorithms such as support vector machines, decision trees, random forest, and linear regression are applied to predict crime rates within clustered regions. The performance of the optimized models is evaluated using standard metrics including accuracy, precision, recall, F1-score, mean absolute error, and root mean square error. Experimental analysis demonstrates that the hybrid K-means and machine learning approach outperforms conventional models trained on unclustered data, highlighting its effectiveness in handling non-linearity and data imbalance. The results indicate that clustering-based optimization significantly enhances crime rate prediction and provides valuable insights for efficient resource allocation and crime prevention. This study contributes to the development of intelligent, data-driven crime prediction frameworks for safer and smarter urban environments.