Penerapan Support Vector Machine dan Random Forest Classifier Untuk Klasifikasi Tingkat Obesitas

Obesity has come become an increasingly concerning global health issue, with 2.5 billion adults experiencing overweight and 890 million identified as obese in 2022. This study aims to develop and compare obesity level classification models using Support Vector Machine (SVM) and Random Forest algorithms, as well as analyze factors influencing obesity. The research utilizes a public dataset comprising 1610 records with 15 variabel encompassing demographic characteristics, family factors, dietary patterns, and lifestyle habits. The research methodology include data preprocessing, dataset splitting with a 70:30 ratio for training and testing, and performance evaluating using accuracy, precision, recall and f1-score metrics. Results indicate that Random Forest demonstrates superior performance with 94% accuracy, a 3% improvement aver SVM which achived 91.01% accuracy. Random Forest showed better consistency in classifying all catagories, particularly achieving optimal results for class 4 with 100% precision and 99% recall. Factor analysisi reveals thet lifestyle and dietary patterns have significant influences on obesity levels. The developed model can be implemented as a supportive tool in healtcare systems to accurately predict and classify obesity levels, enabling more tergeted symtems to accurately predict and classify obesity levels, enabling more tergeted interventions based on identified risk factors.