Forecasting Army Reserve Officer Training Corps Commissions Using Machine Learning Techniques

Abstract

The U.S. Army Cadet Command (USACC) produces about twothirds of the Army officer corps in any given year. The command currently forecasts commissions 24 months before the completion of a fiscal year despite limited ability to influence production within this timeframe. Consequently, USACC must make recruiting decisions to shape its cohorts at least six months before current forecasts begin. This study explores the use of statistical machine learning models to forecast the number of currently enrolled cadets who will commission in a cohort nearly three years out. The developed forecasts can be used to determine the number of new cadets USACC must recruit to accomplish future recruiting missions. We find that a machine learning model can identify predictors that increase or decrease the likelihood of commissioning, and offer insight related to scholarship and contracting policies based on model outputs.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jun 01, 2019
Accession Number
AD1080267

Entities

People

  • Daniel R. Hudalla

Organizations

  • Naval Postgraduate School

Tags

Communities of Interest

  • Autonomy
  • Biomedical

DTIC Thesaurus Topics

  • Attrition
  • Data Science
  • Data Sets
  • Databases
  • Doctrine
  • Education
  • Information Science
  • Machine Learning
  • Military Science
  • Personnel Management
  • Public Policy
  • Random Variables
  • Reserve Officer Training Corps
  • Statistical Analysis
  • Students
  • Training
  • United States

Readers

  • Military Leadership and Professional Education.
  • Naval Personnel Management

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks