Machine Learning Credit Risk Prediction Pipeline

Create an end-to-end machine learning pipeline for predicting credit default probabilities using scikit-learn and TensorFlow. The system must preprocess financial datasets, handle class imbalance, implement advanced feature engineering techniques, and produce explainable AI models. Include automated hyperparameter tuning, cross-validation strategies, and a comprehensive reporting mechanism that translates model predictions into actionable risk insights for financial decision-makers.