Strategies of deep learning for crime forecasting in multiple regions
Guardado en:
| Autores: | , |
|---|---|
| Formáid: | artículo original |
| Stádas: | Versión publicada |
| Fecha de Publicación: | 2026 |
| Cur Síos: | 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. |
| País: | Portal de Revistas TEC |
| Institiúid: | Instituto Tecnológico de Costa Rica |
| Repositorio: | Portal de Revistas TEC |
| Teanga: | Español |
| OAI Identifier: | oai:ojs.pkp.sfu.ca:article/8494 |
| Rochtain Ar Líne: | https://revistas.tec.ac.cr/index.php/tec_marcha/article/view/8494 |
| Palabra clave: | Pronóstico del crimen ajuste de modelos aprendizaje profundo inteligencia artificial Crime forecasting fine-tuned models deep learning artificial intelligence |