DATA-DRIVEN CRIME ANALYSIS AND PREDICTION USING MACHINE LEARNING

Abstract

Crime analysis has become an essential tool for law enforcement agencies and public safety organizations. With the rapid growth of urban populations and digital records, large volumes of crime data are generated daily. This paper presents a data-driven crime analytics and prediction system designed to assist both citizens and law enforcement authorities. The proposed system integrates data preprocessing, exploratory analysis, hotspot detection, and machine learning-based crime prediction within an interactive web platform. Crime datasets uploaded in various formats are automatically processed, cleaned, and analysed to extract meaningful insights. Visualization techniques are used to display crime trends, hotspot areas, and risk indicators. Machine learning models such as Random Forest, Logistic Regression, and time-series forecasting methods are used to predict crime occurrences and future crime trends. The system provides two interfaces: a public portal for crime awareness and safety recommendations, and an administrative dashboard for law enforcement decision support. Experimental results demonstrate that the proposed framework can effectively analyze crime patterns and provide predictive insights that assist authorities in proactive policing strategies.