Citation

Ferrari S, Baumgartner KC, Palermo GB, Bruzzone R, Strano M. IEEE Contr. Syst. Mag. 2008; 28(4): 65-77.

Copyright

(Copyright © 2008, IEEE Control Systems Society, Institute of Electrical and Electronics Engineers)

DOI

10.1109/MCS.2008.924037

PMID

unavailable

Abstract

A knowledge-based system approach in deriving empirical models of offender behavior in criminal psychology has been presented. The approach requires the inputs of psychological characteristics and environmental variables that reflect the mental state of the offender and the circumstances of the crime. This is possible through the study of neural network and Bayesian network models of criminal behavior from databases of single-victim homicides. Such models are utilized to estimate the profile of the offender, in addition to a psychobehavioral portrait that helps detectives select and interrogate suspects. Such a behavioral model yields a mathematical representation of the multidimensional interdependencies between variables that determine offender behavior.