Crime Prediction and Patrol Optimization: H2O Auto ML-Based Recidivism Forecasting, Crime Classification, and Deployment Strategy using Linear Programming

View at publisher Summarise with VIRAT
Authors: Seema Redekar, Sanika Gidye, Aditya Rangari, Arya Angane
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Effective law enforcement along with strategic deployment of the patrol units and accurate forecasting of crimes is crucial. This paper therefore presents a data-driven approach integrating LP for optimal deployment of patrol units with H2O AutoML for predictive analysis, which enables effective crime prevention. The LP model ensures the effective allocation of patrol units as determined by crime intensity and spread and resources, hence ensuring short response times and reduced risk of loss to the public. Meanwhile, H2O AutoML predicts actual crime occurrence and probabilities of recidivism, which allows law enforcement to be proactive in high-risk areas or newly emerging crime trends. Through machine learning methods, our system processes crime history data, detects trends, and maps crime hotspots on geospatial maps. The results demonstrate that integrating LP-based patrol optimization with computerized crime prediction dramatically enhances the efficiency of law enforcement. Implemented as a scalable and flexible framework, this solution is a strong data-based method for preventing crime in different city environments

Fields of Study & Subjects

Crime Patterns and Interventions Data Mining and Machine Learning Applications Criminal Justice and Penology

Publication details

Journal / source
CRC Press eBooks
Pages
162–168
Publication date
25 March 2026
Language
EN
Document type
Book chapter