Credit Default Prediction Machine Learning Pipeline

Create a comprehensive machine learning pipeline using TensorFlow.js for predicting credit default probabilities. Design a system that preprocesses financial datasets, handles feature engineering, trains multiple classification models (logistic regression, random forest), and generates interpretable risk scores. Implement cross-validation techniques, model performance metrics, and a React-based interface for displaying prediction confidence intervals and feature importance.