This article proposes a Bayesian model-based clustering approach for criminal events.
Statistical clustering of criminal events can be used by crime analysts to create lists of potential suspects for an unsolved crime to identify groups of crimes that may have been committed by the same individuals or group of individuals, for offender profiling and for predicting future events. The authors approach is semi-supervised, because the offender is known for a subset of the events, and utilizes spatio-temporal crime locations as well as crime features describing the offender's modus operandi. The hierarchical model naturally handles complex features that are often seen in crime data, including missing data, interval-censored event times and a mix of discrete and continuous variables. In addition, the proposed Bayesian model produces posterior clustering probabilities which allow analysts to act on model output only as warranted. This approach is illustrated by using a large data set of burglaries in 2009-2010 in Baltimore County, Maryland. (Publisher abstract modified)
Downloads
Similar Publications
- Reassessing the Breadth of the Protective Benefit of Immigrant Neighborhoods: A Multilevel Analysis of Violence Risk by Race, Ethnicity, and Labor Market Stratification
- A ROC-based Approximate Bayesian Computation Algorithm for Model Selection: Application to Fingerprint Comparisons in Forensic Science
- Examining Ethno-Racial Related Differences in Child Molester Typology: An MTC:CM3 Approach