PREDICTING CRIMINAL RECIDIVISM - A COMPARISON OF NEURAL-NETWORK MODELS WITH STATISTICAL-METHODS

This article applies neural network and conventional statistical model s to predicting criminal recidivism. While having promising properties for predicting recidivism, the network models do not exhibit any adva ntage over the other methods in an application on a well-known data se t, Analysis suggests that currently available prediction variables hav e limited information content for discriminating recidivists, regardle ss of the models or methods used.