Count Time Series Forecasting Applied to Crime Series in a Multi-Neighborhood Urban Region

• Boanerges T. de Almeida UECE https://orcid.org/0009-0001-6134-0610
• Abraao H. S. Rosal UECE https://orcid.org/0000-0001-7504-9704
• E. B. Maia UECE https://orcid.org/0000-0002-4983-1724

DOI: https://doi.org/10.5753/eniac.2026.30773

Resumo

This article addresses count time series forecasting applied to crime series in Fortaleza City. It involves modeling and forecasting series whose observations are non-negative integer counts, generally with overdispersion and excess zeros. This regime violates the Gaussian assumptions of classical time series models (ARIMA, exponential smoothing), which motivates a specific family of methods. The series of criminal occurrences in Fortaleza (2022–2025) is divided into two contrasting targets: CVLI (Intentional Lethal Violent Crimes), a rare and sparse event; and CVP (Violent Crimes against Property), a frequent and strongly overdispersed event. Prediction experiments with various counting models, namely Poisson/Negative Binomial Regression, INGARCH, INAR, show that they outperform naive baselines. The marginal gain attributable to the autoregressive structure is also investigated.

Referências

Adepeju, M., Rosser, G., and Cheng, T. (2016). Novel evaluation metrics for sparse spatio-temporal point process hotspot predictions-a crime case study. International Journal of Geographical Information Science, 30(11):2133–2154.

Al-Osh, M. A. and Alzaid, A. A. (1987). First-order integer-valued autoregressive (inar (1)) process. Journal of Time Series Analysis, 8(3):261–275.

Andrade, Y., Teodoro, G., Ferreira, R., and Rocha, L. (2025). Um framework para avaliação de modelos de previsão espaço-temporal de crimes. Revista Eletrônica de Iniciação Científica em Computação, 23:136–142.

Butt, U. M., Letchmunan, S., Hassan, F. H., Ali, M., Baqir, A., and Sherazi, H. H. R. (2020). Spatio-temporal crime hotspot detection and prediction: a systematic literature review. IEEE access, 8:166553–166574.

Cameron, A. and Trivedi, P. (1998). Regression analysis of count data. Econometric Society Monographs.

Czado, C., Gneiting, T., and Held, L. (2009). Predictive model assessment for count data. Biometrics, 65(4):1254–1261.

Diebold, F. X. and Mariano, R. S. (1995). Comparing predictive accuracy. Journal of Business & Economic Statistics, 13(3):253–263.

Fokianos, K., Rahbek, A., and Tjøstheim, D. (2009). Poisson autoregression. Journal of the American Statistical Association, 104(488):1430–1439.

Hyndman, R. J. and Koehler, A. B. (2006). Another look at measures of forecast accuracy. International journal of forecasting, 22(4):679–688.

Kolassa, S. (2016). Evaluating predictive count data distributions in retail sales forecasting. International Journal of Forecasting, 32(3):788–803.

Lambert, D. (1992). Zero-inflated poisson regression, with an application to defects in manufacturing. Technometrics, 34(1):1–14.

Ljung, G. M. and Box, G. E. P. (1978). On a measure of lack of fit in time series models. Biometrika, 65(2):297–303.

MacCullagh, P. and Nelder, J. (1989). Generalized linear models. Monographs on statistics and applied probability (37).

Mohler, G. O., Short, M. B., Brantingham, P. J., Schoenberg, F. P., and Tita, G. E. (2011). Self-exciting point process modeling of crime. Journal of the american statistical association, 106(493):100–108.

Mullahy, J. (1986). Specification and testing of some modified count data models. Journal of econometrics, 33(3):341–365.

Villegas, Á. J. R., Pabón, J. S. M., Rubio, M. D., Quintero, S., Vargas, J. G., and García, H. (2022). Spatio temporal sparsity in homicide prediction models. IEEE Access, 10:14359–14367.

Yu, G., Zou, J., Hu, X., Aviles-Rivero, A. I., Qin, J., and Wang, S. (2024). Revitalizing multivariate time series forecasting: Learnable decomposition with inter-series dependencies and intra-series variations modeling. arXiv preprint arXiv:2402.12694.