Machine learning regression model predicting 1985 automobile prices. Lasso model achieves 91.7% R² with superior generalization over XGBoost. Handles extreme multicollinearity (VIF 16,676→8.36), data leakage detection, and outlier treatment through PCA and domain-driven feature engineering.
python data-science machine-learning scikit-learn jupyter-notebook cross-validation regression pandas statistical-analysis xgboost pca data-analysis feature-engineering hyperparameter-tuning multicollinearity lasso-regression model-comparison model-interpretability uci-dataset automobile-pricing-algorithm
-
Updated
Nov 9, 2025 - Jupyter Notebook