Predicting Drinking Behaviour based on Physiological Signals using Ensemble Boosting Algorithms

This paper focuses on Machine learning (ML) classifier-based analysis which offers a promising approach to classify drinking behavior. This study compared four boosting algorithms (Gradient Boosting (GB), Extreme Gradient Boosting (XGB), Adaptive Boosting (ADB), and Histogram Boosting (HB)) for classifying drinkers and non-drinkers using a dataset with individual characteristics and physiological signals. All algorithms achieved good accuracy, with GB and HB showing the best balance between training and testing performance. Additionally, GB, XGB, and HB achieved a high AUC (0.82), demonstrating their effectiveness in classifying drinking behavior. This study highlights the potential of ML for analyzing physiological body signals to classify drinking behavior.