Spatio-temporal Crime Analysis of Bangladesh using Machine Learning Models

DOI

10.2991/978-94-6239-664-7_67How to use a DOI?

Keywords

Bangladesh; spatio-temporal analysis; crime analysis; crime prediction; crime forecasting; data-driven policing; predictive policing; FBProphet; SARIMAX; random forest

Abstract

This study presents a comprehensive spatio-temporal analysis and forecasting of crime trends in Bangladesh using monthly police data from January 2019 to June 2024. This research analyzes five crime types: theft, burglary, robbery, narcotics-related offences, and genderbased violence (GBV) across ten regions, including two major metropolitan cities and eight divisions. Descriptive and exploratory analyses reveal spatial and temporal crime patterns, followed by a comparative evaluation of five forecasting models: Random Forest, XGBoost, SARIMAX, FBProphet, and LSTM. Model performance is assessed using Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE). Results show that model performance varies by crime type. SARIMAX demonstrated strong forecasting capabilities for stationary, auto-correlated data (e.g., theft), while FBProphet performed well for crimes with strong seasonality, such as narcotics-related offences. Random Forest demonstrated strong predictive capability for non-stationary data or stationary data with no autocorrelation, particularly for GBV, whereas LSTM (long short-term memory) underperformed, likely due to the limited dataset size. The comparative performance of the models was further validated by aggregating data in major geographic regions, where SARIMAX, FBProphet, and Random Forest maintained their superiority for theft, narcotics, and GBV, respectively. The study can be used as directives to the policy makers to have detailed knowledge on the spatiotemporal dynamics of crime and choose the right ML models for crime prediction for informed decision making to control future occurrences of crime.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Khaleda Begum
AU  - Md Zamilur Rahman
PY  - 2026
DA  - 2026/06/08
TI  - Spatio-temporal Crime Analysis of Bangladesh using Machine Learning Models
BT  - Proceedings of the International Conference on Intelligent Data Analysis and Applications (IDAA 2025)
PB  - Atlantis Press
SP  - 977
EP  - 993
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-664-7_67
DO  - 10.2991/978-94-6239-664-7_67
ID  - Begum2026
ER  -