Search Results
Title: Predicting Life-Course Persistent Offending Using Machine Learning
Resulting in 1 citation.
| 1. |
Oh, Gyeongseok |
|
Predicting Life-Course Persistent Offending Using Machine Learning Ph.D. Dissertation, Department of Criminal Justice and Criminology, Sam Houston State University, 2021 Cohort(s): NLSY97 Publisher: ProQuest Dissertations & Theses (PQDT) Keyword(s): Crime; Criminal Justice System; Life Course; Modeling, Latent Class Analysis/Latent Transition Analysis Permission to reprint the abstract has not been received from the publisher. The current study investigated the predictive ability of Life-Course-Persistent (LCP) offenders using Machine Learning techniques. Drawing on the National Longitudinal Survey of Youth 1997, LCP and adolescent limited offenders are identified by the latent class growth analysis. Using seven types of Machine Learning techniques, the LCP offenders are predicted by risk factors verified by previous empirical studies. The results of predictive modeling reveal that the Machine Learning-based prediction of LCP offenders significantly outperforms the conventional parametric statistical analysis, logistic regression. Most of all, the predictive ability of Random Forests and Deep Learning model show a more effective forecasting ability than other Machine Learning- based modeling and logistic regression analysis. |
|
|
Bibliography Citation
Oh, Gyeongseok. Predicting Life-Course Persistent Offending Using Machine Learning. Ph.D. Dissertation, Department of Criminal Justice and Criminology, Sam Houston State University, 2021. |