This project predicts credit scores ('Good', 'Standard', 'Poor') using a streamlined ML pipeline. It includes data extraction, cleaning, and preprocessing. Key techniques are Mutual Information for feature selection, PCA for dimensionality reduction, and XGBoost for accurate and efficient model training, ensuring reliable and robust predictions.
python pipeline numpy pandas feature-selection pca dimensionality-reduction sqlite3 mutual-information classification-report imblearn joblib streamlit xgboost-classifier roc-auc-curve random-search-optimization
-
Updated
Sep 5, 2024 - HTML