Identifying Risk Factors of Substance Use in Finland: A Machine Learning Approach

Background: This study explores the dynamics of substance use in Finland, employing Artificial Intelligence (AI) and Machine Learning techniques to identify key risk factors and predict patterns of drug use. Objectives: Using data from the Finnish National Drug Survey and applying 15 feature selection methodologies, this study identifies 23 risk factors associated with the five most used illicit substances: cannabis, ecstasy, amphetamines, cocaine, and non-medical prescription drugs. Results: The most common risk factors include the consumption of e-cigarettes, free and purchased drug offers, and drug-related health problems. This study explores the application of the Bidirectional Long Short-Term Memory (BiLSTM) model for predicting substance use across, demonstrating promising accuracy in this context. Conclusions: By integrating AI methodologies with epidemiological data, this research offers valuable insights for targeted prevention strategies and policy interventions. The findings underscore the complexity of substance use behaviors and highlight the potential of predictive analytics in public health research, aiming to inform and enhance prevention efforts in Finland.