Machine Learning Analysis of the Alcohol and Drug Management Information Tracking System
Abstract
Substance abuse remains a significant problem among Sailors within the U.S. Excessive use of alcohol and drugs can be detrimental to a Sailor's health, safety, and naval service. The U.S. Navy recognizes that substance abuse is preventable and treatable and aims to eliminate these destructive behaviors through continuous training, intervention, and treatment. However, the need to address alcohol and drug abuse using an alternative approach arises as Sailors continue to exhibit these destructive behaviors. We propose the use of supervised machine-learning methods complemented by social network analysis to explore alcohol- or drug - incident data from the Alcohol and Drug Management Information Tracking System (ADMITS). Under our method, we build prediction, classification, forecasting, and generalized network autoregressive (GNAR) models to provide insights on the correlations between Sailors' backgrounds and their alcohol- or drug-related incident information. We utilize performance assessments for regression, classification, and time-dependent data to measure the accuracy of our models and to identify which perform best. Our results strongly demonstrate that the use of supervised machine-learning methods provide an accurate and effective approach to modeling and further understanding of the different aspects of substance abuse among Sailors.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 2021
- Accession Number
- AD1164205
Entities
People
- Vanny Mae O. Angeles
Organizations
- Naval Postgraduate School