Hybrid Machine Learning Approach for Crime Prediction and Forecasting: Integrating K-means Clustering and LSTM Networks

References

[1] G. Mohler, "Homicide and gun crime prediction in Chicago using marked point process hotspot maps," Int. J. Forecasting, July 2022, vol. 30, no. 3, pp. 491–497.

[2] 'Natural language processing and e-government: Extracting reusable crime report information,' in Proc. IEEE Int. Conf. Inf. Reuse Integr., Las Vegas, IL, USA, Aug. 2020, pp. 221–226. A. Iriberri and G. Leroy.

[3] In Proc. IEEE Int. Conf. Intell. Secur. Informat., Vancouver, BC, Canada, May 2021, pp. 19–24, V. Pinheiro, V. Furtado, T. Pequeno, and D. Nogueira, "Natural language processing based on semantic inferentialism for extracting crime information from text."

[4] P. J. Brantingham, S. J. Osher, J. Xin, B. Wang, P. Yin, A. L. Bertozzi, and "Deep learning for real-time crime forecasting and its ternarization," Chin. Ann. Math., B, vol. 40, no. 6, pp. 949–966, Nov. 2020.

[5] S. Chackravarthy, S. Schmitt, and L. Yang, "Deep learning-based intelligent crime anomaly detection in smart cities," in Proceedings of IEEE 4th International Conference on Collaboration Internet Comput. (CIC), Philadelphia, PA, USA, October 2021, pp. 399–404.

[6] "Prediction of crime occurrence from multimodal data using deep learning," by H.-W. and H.-B. Kang Art. no. e0176244, PLoS ONE, vol. 12, no. 4, April 2019.

[7] As stated in Social Media Strategy in Policing (Security Informatics and Law Enforcement), B. Akhgar, P. S. Bayeri, and G. Leventakis, Eds. Cham, Switzerland: Springer, 2019, pp. 177–195, A. Fidow, M. Hassan, M. Imran, X. Cheng, C. Petridis, and C. Sule, "Proposing a hybrid approach mobile apps with big data analysis to report and prevent crimes."

[8] Brantingham, M. Valasik, and G. O. Mohler, P. J. "Does arrest bias result from predictive policing? A randomized controlled trial's findings, Statist. Public Policy, vol. 5, no. 1, pp. 1–6, January 2020.

[9] Adv. Sci. Technol. Lett., vol. 90, pp. 90–92, Dec. 2021; A. Nasridinov and Y.-H. Park, "A study on performance evaluation of machine learning algorithms for crime dataset,"

[10] "Forecasting crime with deep learning," by A. Stec and D. Klabjan arXiv:1806.01486, 2022. [Online]. http://arxiv.org/abs/1806.01486 is accessible.

@article{1707748,
      author = {Karthikeyan M, Naveen Kumar N, Maruthupandi M},
      title = {Hybrid Machine Learning Approach for Crime Prediction and Forecasting: Integrating K-means Clustering and LSTM Networks},
      journal = {Iconic Research And Engineering Journals},
      year = {2025},
      volume = {8},
      number = {10},
      pages = {99-105},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1707748.pdf},
      abstract = {For law enforcement organizations to maintain public safety and effectively deploy resources, crime prediction and forecasting are essential duties. In order to produce precise crime forecasts, this study offers a novel strategy that blends machine learning and deep learning methodologies. In order to take use of their complementing strengths in recognizing crime patterns and modeling sequential data, we suggest combining k-means clustering with Long Short-Term Memory (LSTM) networks. The first step in the procedure is gathering and preparing various crime data, such as demographic data and crime reports. The chronological, spatial, and social components of the extracted variables offer important insights into criminal incidents. Similar crime data points are organized into clusters using k-means clustering, which indicates unique crime patterns or hotspots. After these clusters are converted into sequential data, LSTM models can use historical crime sequences to forecast future crime patterns. In terms of crime prediction and forecasting, the hybrid approach shows encouraging results, giving law enforcement useful information to stop and efficiently address criminal activity. However, while using sensitive crime-related data, ethical issues and data protection laws should be strictly followed.},
      keywords = {Crime Prediction, Machine Learning, Analysis and Forecast, Metropolitan Safety},
      month = {April},
  }