Module 1: Crime Analysis

This weeks lab provided insight into different methods used with crime hotspot analysis.  and explored the differences between global and local approaches. 

The lecture defines global clustering as a pattern which produces a single statistic with confidence intervals and local clustering which produces a hotspot.  Some of the more notable methods for determining local clusters are grid-based mapping, local Moran's I, kernel density, Gi*, and Nearset Neighbor Hierarchical (NNH). These method combined data aggregation provide valuable insights for decision makers to improve their community.

Kernel Density Hotspot

Geographic Boundary Hotspot

Some examples of hotspot mapping found in the lab is shown above, which show a kernel density of assault and choropleth map of burglary rate. Below are examples of grid-based, kernel density, and local Moran's I hotspot mapping.

Finally, the lab concluded with creating a table which provide a matrix (not shown) to evaluate the different methods. Utilizing the information found in the map and the matrix the kernel density and local Moran's I offered the greatest crime hotspot insight, where the the local Moran's I performed slightly better.