Abstract

Obesity is a global health issue with a continuously rising prevalence and serves as a major risk factor for various chronic diseases. Early prediction of obesity levels is crucial to mitigate its negative impacts. The application of Machine Learning methods presents a potential solution to this problem. However, the primary challenge lies in selecting a highly accurate machine learning method. Thus, it is necessary to develop a reliable classification model using high-quality data and appropriate algorithms to support decision-making in healthcare, particularly in obesity level prediction. This study aims to identify the most accurate model among decision tree-based machine learning methods for predicting obesity levels. The methods examined include Gradient Boosting, Random Forest, and C4.5. The findings indicate that the Gradient Boosting algorithm achieved the highest accuracy of $96.31 \%$, followed by Random Forest at $94.59 \%$, and C 4.5 at $93.19 \%$. These results suggest that Gradient Boosting has potential as a decision-support tool in healthcare for early obesity detection. Its implementation could assist medical professionals in designing faster, more precise, and targeted interventions, thereby minimizing the adverse effects of obesity more effectively.

Concepts :

Access to Document

Citations by Year