Strategies of deep learning for crime forecasting in multiple regions
Article Sidebar
Main Article Content
Abstract
This study compares the crime prediction performance across 83 regions of fine-tuned pre-trained models versus models trained from scratch, using different strategies. The fine-tuned Lag-Llama model, using a strategy of training of a unique model that can predict any of the 83 regions was the best for monthly predictions, while the fine-tuned Lag-Llama using the strategy of training by groups of time series created with the k-means method was the best for the daily predictions. Apparently, the clustering training strategy allows the Lag-Llama to make a better fine-tuned for time series with characteristics that make them less predictable, such as nonlinearity and variability. Even though the Lag-Llama showed the best results at the general level, it is not the best model to make crime predictions for every region. There are models more suitable for some regions. Therefore, it is advisable to implement more than one model in a crime forecasting system.
Article Details
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Los autores conservan los derechos de autor y ceden a la revista el derecho de la primera publicación y pueda editarlo, reproducirlo, distribuirlo, exhibirlo y comunicarlo en el país y en el extranjero mediante medios impresos y electrónicos. Asimismo, asumen el compromiso sobre cualquier litigio o reclamación relacionada con derechos de propiedad intelectual, exonerando de responsabilidad a la Editorial Tecnológica de Costa Rica. Además, se establece que los autores pueden realizar otros acuerdos contractuales independientes y adicionales para la distribución no exclusiva de la versión del artículo publicado en esta revista (p. ej., incluirlo en un repositorio institucional o publicarlo en un libro) siempre que indiquen claramente que el trabajo se publicó por primera vez en esta revista.
References
[1] A. Meijer and M. Wessels, “Predictive Policing: Review of Benefits and Drawbacks,” International Journal of Public Administration, vol. 42, no. 12, pp. 1031–1039, Feb. 2019, doi: 10.1080/01900692.2019.1575664.
[2] U. M. Butt, S. Letchmunan, F. H. Hassan, and T. W. Koh, “Hybrid of deep learning and exponential smoothing for enhancing crime forecasting accuracy,” PLOS ONE, vol. 17, no. 9, p. e0274172, Sep. 2022, doi: 10.1371/journal.pone.0274172.
[3] K. Rasul, A. Ashok, A. R . Williams, A.Khorasani, G.Adamopoulos, R. Bhagwatkar, I.Rish, Lag-llama: Towards foundation models for time series forecasting, 2023. Available: arXiv preprint arXiv:2310.08278.
https://doi.org/10.48550/arXiv.2310.08278
[4] A. Garza, A, M. Mergenthaler-Canseco. TimeGPT-1, 2023. Available: arXiv preprint arXiv:2310.03589.
https://doi.org/10.48550/arXiv.2310.03589
[5] G. Woo, C. Liu, A. Kumar, C. Xiong, S. Savarese, and D. Sahoo, “Unified training of universal time series forecasting transformers,” *arXiv*, 2024. Available: https://doi.org/10.48550/arXiv.2402.02592
[6] M. Solís and L.-A. Calvo-Valverde, “Deep Learning for Crime Forecasting of Multiple Regions, Considering Spatial–Temporal Correlations between Regions,” ITISE 2024, p. 4, Jun. 2024, doi: 10.3390/engproc2024068004.
[7] V. Mandalapu, L. Elluri, P. Vyas, and N. Roy, “Crime prediction using machine learning and deep learning: A systematic review and future directions,” IEEE Access, vol. 11, pp. 60153–60170, 2023, doi: 10.1109/ACCESS.2023.3286344.
[8] P. Stalidis, T. Semertzidis, and P. Daras, “Examining deep learning architectures for crime classification and prediction,” Forecasting, vol. 3, no. 4, pp. 741-762, 2021, doi: 10.3390/forecast3040046.
[9] V. Madhu, “Crime Rate Prediction Using Time Series Forecasting,” Master’s thesis, National College of Ireland, 2023.
[10] J. Austin and R. Rosenfeld, “Forecasting US Crime Rates and the Impact of Reductions in Imprisonment: 1960–2025,” CrimRxiv, 2023. [Online]. Available: https://doi.org/10.21428/cb6ab371.395a314b
[11] H. K. Sharma and R. M. Tailor, “Exploring the Potential of AI for Urban Crime Prediction in India: A Case Study of Indian Metropolitan Cities,” in 2023 6th International Conference on Contemporary Computing and Informatics (IC3I), 2023, pp. (if known), doi: 10.1109/IC3I59117.2023.10398043.
[12] K. Hu, L. Li, J. Liu, and D. Sun, “DuroNet,” ACM Transactions on Internet Technology, vol. 21, no. 1, pp. 1–24, 2021, doi: 10.1145/3432249.
[13] S. Sharmin, F. I. Alam, A. Das, and R. Uddin, “An Investigation into Crime Forecast Using Auto ARIMA and Stacked LSTM,” in 2022 International Conference on Innovations in Science, Engineering and Technology (ICISET), 2022, pp. (if known), doi: 10.1109/ICISET54810.2022.9775862.
[14] N. Ibrahim, S. Wang, and B. Zhao, “Spatiotemporal crime hotspots analysis and crime occurrence prediction,” in Advanced Data Mining and Applications: 15th International Conference, ADMA 2019, Dalian, China, November 21–23, 2019, Proceedings 15. Springer International Publishing, 2019, pp. 579-588.
[15] M. A. Cruz-Nájera et al., “Short Time Series Forecasting: Recommended Methods and Techniques,” Symmetry, vol. 14, no. 6, pp. 1231-1243, 2022, doi: 10.3390/sym14061231.
[16] J. Alghamdi and Z. Huang, “Modeling daily crime events prediction using seq2seq architecture,” in Databases Theory and Applications: 32nd Australasian Database Conference, ADC 2021, Dunedin, New Zealand, January 29–February 5, 2021, Proceedings 32. Springer International Publishing, 2021, pp. 192-203.
[17] Q. Dong, Y. Li, Z. Zheng, X. Wang, and G. Li, “ST3DNetCrime: Improved ST-3DNet Model for Crime Prediction at Fine Spatial Temporal Scales,” ISPRS International Journal of Geo-Information, vol. 11, no. 10, p. 529, 2022, doi: 10.3390/ijgi11100529.
[18] W. Safat, S. Asghar, and S. A. Gillani, “Empirical Analysis for Crime Prediction and Forecasting Using Machine Learning and Deep Learning Techniques,” IEEE Access, vol. 9, pp. 70080-70094, 2021, doi: 10.1109/ACCESS.2021.3078117.
[19] A. Stec and D. Klabjan, “Forecasting crime with deep learning,” arXiv, 2018. [Online]. Available: https://doi.org/10.48550/arXiv.1806.01486
[20] M. Ashby, “Forecasting crime trends to support police strategic decision making,” CrimRxiv, 2023. [Online]. Available: https://doi.org/10.21428/cb6ab371.8c79f146
[21] Y. Nie, N. H. Nguyen, P. Sinthong, and J. Kalagnanam, “A time series is worth 64 words: Long-term forecasting with transformers,” arXiv, 2022. [Online]. Available: https://doi.org/10.48550/arXiv.2211.14730
[22] Y. Kang, R. J. Hyndman, and K. Smith-Miles, “Visualising forecasting algorithm performance using time series instance spaces,” International Journal of Forecasting, vol. 33, no. 2, pp. 345–358, Apr. 2017, doi: 10.1016/j.ijforecast.2016.09.004.
[23] X. Wang, K. Smith, and R. Hyndman, “Characteristic-Based Clustering for Time Series Data,” Data Mining and Knowledge Discovery, vol. 13, no. 3, pp. 335–364, May 2006, doi: 10.1007/s10618-005-0039-x.
[24] A. F. Ansari et al., “Chronos: Learning the language of time series,” arXiv, 2024. [Online]. Available: https://doi.org/10.48550/arXiv.2403.07815
[25] M. Solís and L. A. Calvo-Valverde, “A Proposal of Transfer Learning for Monthly Macroeconomic Time Series Forecast,” Eng. Proc., vol. 39, no. 1, p. 58, 2023.