This project applies machine learning techniques to predict obesity levels based on lifestyle, dietary, and demographic factors.
The motivation is to explore how supervised learning models can help identify obesity risks early and support health awareness.
Key highlights:
- Preprocessing real-world obesity data with missing values
- Using KNN imputation for handling missing data
- Comparing classifiers such as Logistic Regression, Decision Trees, KNN, and Random Forest
- Selecting the best-performing model (Random Forest) for classification
- Saving and reusing the trained model for predictions
By understanding the factors that contribute most to obesity, this project also demonstrates how data-driven approaches can be used in public health and preventive medicine.
- Data preprocessing & missing value handling
- Feature selection & transformation
- Model training & hyperparameter tuning
- Model evaluation with performance metrics
- Final Random Forest model deployment
...