Analysis of Criminal Landscape by Utilizing Statistical Analysis and Deep Learning Techniques

Abstract

This research intends to provide law enforcement organizations with a deep learning model that uses trends in past crimes to forecast future crimes. The study will enable them to dispatch security patrols to the most susceptible regions and take preventive actions. The experimental study demonstrates that the LSTM-based deep learning method beats the conventional ARIMA model. Multiple parameter tuning techniques were examined to create an optimized model, such as different LSTM layers, epochs, and batch sizes. The developed model has a training accuracy of around 90%, while on the test data, the minimum and highest accuracy levels were around 75%.

Keywords:

• Crime prediction
• deep learning
• forecasting
• statistical analysis
• time series

Authors’ contributions

RKM conceived this research and designed experiments; ARA participated in the design and interpretation of the data; RKM performed experiments and analysis; RKM, ARA, JAA, and VM wrote the paper and participated in the revisions of it. All authors read and approved the final manuscript.

Disclosure statement

The authors declare they have no conflicts of interest to report regarding the present study.

Data availability statement

Authors declare that all the data being used in the design and production cum layout of the manuscript is declared in the manuscript.

Additional information

Funding

The authors received no specific funding for this study.