📊 Polynomial Regression with Ridge Regularization

This project demonstrates a complete end-to-end regression modeling workflow following a structured data science process.

Starting from a baseline Linear Regression model, we:

Evaluated performance using cross-validation

Compared results against a mean baseline

Diagnosed underfitting

Introduced Polynomial Features (degree=2) to capture non-linear relationships

Applied Ridge regularization to control variance

Tuned hyperparameters using GridSearchCV

Evaluated final performance on a held-out test set

🔎 Final Model Performance

Test RMSE: ~0.76

Test R²: ~0.93

The final model (Polynomial Features + Ridge Regression) explains 93% of the variance in unseen data, demonstrating strong generalization and proper bias–variance tradeoff management.

🚀 Key Concepts Demonstrated

Bias–variance tradeoff

Cross-validation

Feature engineering (interaction & quadratic terms)

Regularization (L2 / Ridge)

Model evaluation and diagnostics

Proper train/test workflow

This repository illustrates a practical and interview-ready example of structured regression modeling in Python using scikit-learn.