Machine Learning Pipeline

Design a complete ML pipeline for predicting customer churn: data preprocessing with feature engineering, exploratory data analysis template, model selection comparison (logistic regression, random forest, XGBoost, neural network), cross-validation strategy, hyperparameter tuning with Optuna, feature importance analysis, model interpretability with SHAP values, and deployment as REST API