Modeling Recidivism through Bayesian Regression Models and Deep Neural Networks

This study aims to analyze and explore criminal recidivism with different modeling strategies: one based on an explanation of the phenomenon and another based on a prediction task. We compared three common statistical approaches for modeling recidivism: the logistic regression model, the Cox regression model, and the cure rate model. The parameters of these models were estimated from a Bayesian point of view. Additionally, for prediction purposes, we compared the Cox proportional model, a random survival forest, and a deep neural network. To conduct this study, we used a real dataset that corresponds to a cohort of individuals which consisted of men convicted of sexual crimes against women in 1973 in England and Wales. The results show that the logistic regression model tends to give more precise estimations of the probabilities of recidivism both globally and with the subgroups considered, but at the expense of running a model for each moment of the time that is of interest. The cure rate model with a relatively simple distribution, such as Weibull, provides acceptable estimations, and these tend to be better with longer follow-up periods. The Cox regression model can provide the most biased estimations with certain subgroups. The prediction results show the deep neural network’s superiority compared to the Cox proportional model and the random survival forest.

[1]  Balasubramanian Narasimhan,et al.  A scalable discrete-time survival model for neural networks , 2018, PeerJ.

[2]  Herman J. Bierens,et al.  Semi-nonparametric competing risks analysis of recidivism , 2007 .

[3]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[4]  D Faraggi,et al.  A neural network model for survival data. , 1995, Statistics in medicine.

[5]  P. Gendreau,et al.  A META‐ANALYSIS OF THE PREDICTORS OF ADULT OFFENDER RECIDIVISM: WHAT WORKS!* , 1996 .

[6]  Terry L. Schell,et al.  Predicting DUI recidivism: Personality, attitudinal, and behavioral risk factors. , 2006, Drug and alcohol dependence.

[7]  A L Shroyer,et al.  Bayesian-logit model for risk assessment in coronary artery bypass grafting. , 1994, The Annals of thoracic surgery.

[8]  Maher Rizkalla,et al.  Deep learning-based cancer survival prognosis from RNA-seq data: approaches and evaluations , 2020, BMC Medical Genomics.

[9]  B. Francis,et al.  Competing Risks, Persistence, and Desistance in Analyzing Recidivism , 2000 .

[10]  C. Robert,et al.  Bayesian Modeling Using WinBUGS , 2009 .

[11]  A. Piquero Assessing the relationships between gender, chronicity, seriousness, and offense skewness in criminal offending , 2000 .

[12]  Peter Congdon,et al.  Applied Bayesian Modelling , 2003 .

[13]  F. Harrell,et al.  Evaluating the yield of medical tests. , 1982, JAMA.

[14]  Alex Alves Freitas,et al.  Comprehensible classification models: a position paper , 2014, SKDD.

[15]  William Nick Street,et al.  A Neural Network Model for Prognostic Prediction , 1998, ICML.

[16]  David W. Hosmer,et al.  Applied Logistic Regression , 1991 .

[17]  Uri Shaham,et al.  DeepSurv: personalized treatment recommender system using a Cox proportional hazards deep neural network , 2016, BMC Medical Research Methodology.

[18]  R. K. Hanson,et al.  Predicting relapse: a meta-analysis of sexual offender recidivism studies. , 1998, Journal of consulting and clinical psychology.

[19]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[20]  P. Schmidt,et al.  Predicting Criminal Recidivism Using "Split Population" Survival Time Models , 1987 .

[21]  Andrew D. Martin,et al.  MCMCpack: Markov chain Monte Carlo in R , 2011 .

[22]  Susan W. Palocsay,et al.  Predicting criminal recidivism using neural networks , 2000 .

[23]  Hemant Ishwaran,et al.  Random Survival Forests , 2008, Wiley StatsRef: Statistics Reference Online.

[24]  Aj Beck,et al.  Recidivism of prisoners released in 1983 , 1989 .

[25]  D. Sargent,et al.  Comparison of artificial neural networks with other statistical approaches , 2001, Cancer.

[26]  I. Cha,et al.  Deep learning-based survival prediction of oral cancer patients , 2019, Scientific Reports.

[27]  P. Lapuerta,et al.  Comparison of the performance of neural network methods and Cox regression for censored survival data , 2000 .

[28]  Nikolaj Tollenaar,et al.  Optimizing predictive performance of criminal recidivism models using registration data with binary and survival outcomes , 2019, PloS one.

[29]  R. Maller,et al.  Survival Analysis with Long-Term Survivors , 1996 .