PENERAPAN DECISION TREE MENGGUNAKAN ALGORITMA CART DALAM PREDIKSI INDEKS MASSA TUBUH BERBASIS WEB

Abstract

Body Mass Index (BMI) is an important indicator for assessing an individual's nutritional status, but manual BMI measurements are often impractical and time-consuming. This study aims to develop a web-based BMI prediction system using the Classification and Regression Tree (CART) algorithm to facilitate BMI calculation and provide health recommendations. The system measures BMI status based on three main factors: weight, height, and gender, with classification into six categories: Extreme Underweight, Underweight, Normal, Overweight, Obese, and Extreme Obese. The research methodology includes five main stages: data collection, data preparation, CART model training, data splitting, and website design. The CART model was chosen due to its ability to classify data with high accuracy. Model testing was performed using a dataset consisting of 500 samples, with 80% used for training and 20% for testing. Evaluation results show that the model achieved an accuracy of 89%, with other evaluation metrics such as precision, recall, and F1-score indicating good performance. The system is also equipped with an interactive web interface to facilitate users in accessing BMI prediction results in real-time, which is expected to raise public awareness about the importance of monitoring health status.