LEVERAGING MACHINE LEARNING FOR HEALTH IMPACT ANALYSIS OF SOFT DRINK CONSUMPTION USING ENSEMBLE METHODS

Soft drinks, often high in added sugars, have raised significant public health concerns due to their association with adverse health effects such as obesity, type 2 diabetes, and cardiovascular diseases. The aim of this study is to investigate the health impact of soft drink consumption, particularly focusing on the implications of excessive sugar intake. To address this concern, we employed four Ensemble Learning approaches to assess the health risks associated with soft drink consumption, specific Ensemble Learning approaches such as LightGBM, CatBoost, XGBoost, and Random Forest we adopted. Our findings indicate that older individuals may not require as much soft drink consumption, and the general population should limit their daily intake to a single bottle, approximately 35g per day. The study reveals that using Ensemble Learning techniques, including LightGBM, CatBoost, XGBoost, and Random Forest, yielded promising results, with CatBoost emerging as the top-performing model with a model accuracy score of 97%, surpassing the performance of Random Forest and XGBoost algorithms. The research reveals the importance of limiting the consumption of sugary beverages and opting for healthier alternatives to mitigate the risk of adverse health outcomes associated with excessive sugar consumption. Overall, our findings contribute valuable insights into understanding the health implications of soft drink consumption and highlight the efficacy of Ensemble Learning approaches in assessing and addressing public health concerns. This study provides actionable recommendations for individuals and health organizations to promote healthier dietary habits and mitigate the risk of sugar-related health complications, emphasizing the importance of evidence-based interventions in safeguarding public health