Machine Learning in Social Psychology:From Active Addiction to Rehabilitation

Long abstract: Substance use disorders (SUDs) are complex, highly dimensional conditions that are influenced by a variety of personal and environmental factors. For example, conditioned environmental cues can either support the maintenance of a SUD or act as a barrier to recovery. Conversely, stable housing, strong social networks, and other measures of social and personal recovery capital can be supportive and protective of abstinence. Traditional statistical methods may not be sufficient for fully analyzing and interpreting the outcomes of SUDs, given their complexity and the high number of factors involved. In order to better understand these factors, researchers have begun to turn to machine learning (ML) models, which can provide a data-driven approach to studying SUDs by making use of large datasets. Our systematic review of recent literature (2019-2022) on the use of ML in addiction research found that this approach is becoming increasingly popular compared to years prior. ML models are especially useful for analyzing highly dimensional targets and can help us gain a more comprehensive understanding of the factors that influence addiction. In this poster, we summarize the current state of ML in addiction research, discuss its potential value, and highlight gaps and disparities in its use. We also explore potential future applications of this approach, including the possibility of using ML to develop more effective interventions and treatments for SUDs