This project explores feature selection techniques using the Obesity Dataset from the UCI Machine Learning Repository.
- Logistic Regression
- Sequential Forward Selection (SFS)
- Sequential Backward Floating Selection (SBFS)
- Recursive Feature Elimination (RFE)
Obesity Levels Based on Eating Habits and Physical Condition
- pandas
- numpy
- scikit-learn
- mlxtend
- matplotlib
| Model | Accuracy |
|---|---|
| Logistic Regression | 76.60% |
| Sequential Forward Selection | 78.35% |
| Sequential Backward Floating Selection | 78.30% |
| Recursive Feature Elimination | 76.79% |
Sequential Forward Selection (SFS) achieved the highest accuracy of 78.35% while reducing the number of features from 18 to 9.