Psychosocial Factors Predict the Level of Aggression of People with Drug Addiction: A Machine Learning Approach

ABSTRACT

This study aimed to identify the relevant psychosocial factors that can predict the aggression in people with drug addiction. A total of 896 male participants (Mean_age = 38.30 years) completed the survey. Gradient boosting regression, a machine learning algorithm, was used to find the relevant psychosocial variables, such as psychological security, psychological capital, interpersonal trust and alexithymia, that may be significantly related to aggressive behavior. Results showed that the five most important factors in the prediction of aggression are interpersonal trust, psychological security, psychological capital, parental conflict and alexithymia. A high level of interpersonal trust, psychological security and psychological capital can predict a low level of aggression in people with drug addiction, while a high level of parental conflict and alexithymia can predict a high level of aggression. Overall, the findings highlight the need to focus interventions according to the relation between these psychosocial factors and aggression in order to decrease violence.

KEYWORDS:

• Aggressive behavior
• drug addiction
• psychosocial variables
• machine learning

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by, The Ministry of Education of Humanities and Social Science Project [18YJA190012] and Guangdong Province Philosophy and Social Science Foundation [GD19CSH03].