This project aims to develop and compare different classification models for predicting obesity levels based on individual characteristics and lifestyle factors. The dataset contains features such as gender, age, height, weight, family history of overweight, dietary habits, and physical activity levels, among others.
- Data Preprocessing: The dataset underwent extensive preprocessing, including handling missing values, encoding categorical variables, and standardizing numerical features.
- Model Selection: Three classification algorithms were chosen for comparison: K-Nearest Neighbors (KNN), Multi-layer Perceptron (MLP), and Gradient Boosting Machine (GBM).
- Model Training and Evaluation: Each model was trained on the preprocessed dataset and evaluated using performance metrics such as accuracy, precision, recall, and F1-score. Confusion matrices were also generated to visualize classification results.
- Comparison and Analysis: The performance of each model was compared to identify the most effective approach for obesity prediction.
Precision, recall, and F1-score were compared across models, revealing insights into their strengths and weaknesses. Graphical visualizations, including bar plots and confusion matrices, were utilized to present the comparative analysis of model performance.
Based on the results, the MLP classifier demonstrated superior performance for predicting obesity levels in individuals. This project highlights the importance of selecting appropriate classification algorithms and evaluating their performance in healthcare-related prediction tasks.