An analysis of Edmonton's Homicides using Spatial Statistics

Abstract

Prior to 2011, Edmonton had never experienced 40 homicides in a single year. Since then, there have been three separate years to experience 40 or more murders. With this knowledge, it seems prudent to explore any relationships that may exist regarding homicide events. While there has been research done into what types of people are involved in such events, little research can be found on spatial trends associated with homicide locations. Using spatial statistics, it is my goal to explore the possible existence of a spatial relationship between homicide events. More specifically, point pattern analysis will be used to investigate possible clustering of events and areal analysis will be used to check for spatial autocorrelation. If possible, these analyses will be followed by attempts to model measurable covariates that may share a similar spatial relationship with the hopes of determining at risk areas.

Discipline: Mathematics

Faculty Mentor: Dr. Rui Hu