multi-dimension exponent approximation

This repository contains the data and code used to study space-time clustering in large crime datasets using self-exciting point process models. We propose a multi-dimensional Gaussian-type exponent approximation that improves computational efficiency and allows flexible estimation of the conditional intensity function. The method is evaluated through simulations and applied to burglary data from Chicago, where strong space-time clustering patterns are observed. The provided materials support reproducible analysis and can be adapted to other large-scale spatiotemporal event datasets.