Deep Learning for Real Time Crime Forecasting

TL;DR

This paper adapts a deep learning spatio-temporal model to predict crime distribution in Los Angeles, addressing the challenges of crime's low regularity and complex factors for real-time public safety applications.

Contribution

It introduces a two-stage approach with data preprocessing and hierarchical residual convolutional networks for improved crime prediction accuracy.

Findings

High predictive accuracy over half a year in Los Angeles

Effective handling of crime's low regularity in space and time

Enhanced signal extraction through data regularization

Abstract

Accurate real time crime prediction is a fundamental issue for public safety, but remains a challenging problem for the scientific community. Crime occurrences depend on many complex factors. Compared to many predictable events, crime is sparse. At different spatio-temporal scales, crime distributions display dramatically different patterns. These distributions are of very low regularity in both space and time. In this work, we adapt the state-of-the-art deep learning spatio-temporal predictor, ST-ResNet [Zhang et al, AAAI, 2017], to collectively predict crime distribution over the Los Angeles area. Our models are two staged. First, we preprocess the raw crime data. This includes regularization in both space and time to enhance predictable signals. Second, we adapt hierarchical structures of residual convolutional units to train multi-factor crime prediction models. Experiments over a half year period in Los Angeles reveal highly accurate predictive power of our models.