In this lab, we learned to use Crime Stats to analyze crime patterns. We conducted Nearest Neighbor Analysis, Hot Spot Analysis, Kernel Density Analysis and Moran's I analysis to Residential Break and Enter crimes, Commercial Break and Enter crimes, Stolen Vehicles, and Robberies in Ottawa city 2006. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Nearest neighbor hierarchical clustering can indicate crime hotspots. Hierarchical technique is like an inverted tree diagram, in which two or more incidents are first grouped on the basis of some criteria, in this case, close neighbors within 1000m radius. (CrimeStat) The result of this analysis is showing as pink ellipses on map 1. Fuzzy Mode hotspot analysis and nearest neighbor hierarchical clustering analysis reflects similar result. The fuzzy mode frequency dots overlay with the nearest neighbor hierarchical ellipses.

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

kernel density interpolation is a technique for generalizing incident locations to the entire study area. Single kernel density is applied to a distribution of Residential B&E incidents. Map 2 shows the result of this analysis. Red indicates areas with highest density of incidents, and blue shows areas with lowest density of incidents. Dual kernel density is applied to two distributions. To consider two attributes, two files are chosen for analysis. The primary file was Residential B&E incident, and the secondary file was population census. In this case, population is considered in determining density of crime incident. Map 3 shows the result of this analysis. Again, red indicates highest density of residential B&E, blue indicates the lowest density. Comparing the result of single kernel density (map2) and dual kernel density (map3), by adding the population factor, many areas considered as high crime density in single kernel density analysis is indicated as low crime density in dual analysis.

​

​

​