# Public decision support for low population density areas: An imbalance-aware hyper-ensemble for spatio-temporal crime prediction

## TL;DR

This paper introduces an imbalance-aware hyper-ensemble machine learning approach for predicting crime hotspots in low population density areas, significantly improving detection accuracy despite data sparsity.

## Contribution

It presents a novel hyper-ensemble model specifically designed to handle class imbalance in spatio-temporal crime prediction in low-density regions.

## Key findings

- Hit ratio increased from 18.1% to 24.6% for top 5% hotspots.
- Hit ratio increased from 53.1% to 60.4% for top 20% hotspots.
- Model outperforms state-of-the-art predictors in imbalanced settings.

## Abstract

Crime events are known to reveal spatio-temporal patterns, which can be used for predictive modeling and subsequent decision support. While the focus has hitherto been placed on areas with high population density, we address the challenging undertaking of predicting crime hotspots in regions with low population densities and highly unequally-distributed crime.This results in a severe sparsity (i.e., class imbalance) of the outcome variable, which impedes predictive modeling. To alleviate this, we develop machine learning models for spatio-temporal prediction that are specifically adjusted for an imbalanced distribution of the class labels and test them in an actual setting with state-of-the-art predictors (i.e., socio-economic, geographical, temporal, meteorological, and crime variables in fine resolution). The proposed imbalance-aware hyper-ensemble increases the hit ratio considerably from 18.1% to 24.6% when aiming for the top 5% of hotspots, and from 53.1% to 60.4% when aiming for the top 20% of hotspots.