Machine Learning-Based Prediction of Opioid and Stimulant-related Overdose Death Rates

This repository contains the data and R code used to develop, train, optimize, and evaluate machine learning models for overdose fatality prediction. It includes preprocessed monthly overdose mortality data, feature engineering scripts, ARIMA, Random forest, and XGBoost modeling workflows, Bayesian hyperparameter optimization, model evaluation, visualization scripts, and blended model implementation to facilitate reproducibility of the study findings.