Classification trees outperform logistic regression predictions of attrition in the U.S. Marine Corps

The present study compared the performance of machine learning classification models against logistic regression in the context of predicting training attrition from the Delayed Enlistment Program in the United States Marine Corps (UMSC) with scores from the Tailored Adaptive Personality Assessment System (TAPAS). The base-rate of attrition was low which made the model training process difficult, but the random-forest model outperformed logistic regression in predicting cases of attrition in a stratified 50% attrition sample.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Psychology

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2020

Author dc:creator

• Alzate Vanegas, Juan Manuel

Contributors dc:contributor

• Drasgow, Fritz

• Attrition
• turnover
• TAPAS
• logistic regression
• machine learning
• CART
• LASSO
• random forests

Statement dc:rights

• © 2020 Juan Manuel Alzate Vanegas

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/108465

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/108465