🧠 Regularized Regression with Ridge & LASSO (Python)

This project applies regularized regression techniques—Ridge and LASSO—on the Boston Housing dataset to demonstrate their effectiveness in handling multicollinearity and enhancing prediction accuracy.

📊 Dataset

• Source: Boston Housing Dataset
• Target Variable: medv (Median house value)

🧪 Workflow

1. Load and preview the dataset
2. Pre processing data: duplicate check, unvalid and missing value check, handling outlier
3. Split data into:
   • Pre-train: 80%
     • Train 80%
     • Validation 20%
   • Test: 20%
4. Calculate VIF for multicollinearity check
5. Feature scaling using StandardScaler
6. Train Ridge & LASSO models with cross-validation
7. Applied on test set
8. Compare models using RMSE, MAE, MAPE, and R²
9. Choose the best model

📈 Model Performance

Model	Train RMSE	Test RMSE	R² Score
Ridge	4.47	3.23	75.32
LASSO	4.47	3.19	75.25

📁 Files

• Regularized Regression using Python.ipynb: Full code
• Workflow.png: Python code workflow

#DataScience #MachineLearning #Regularization #RidgeRegression #LassoRegression #Modeling #PredictiveAnalytics #MLTips