Machine Learning Feature Extraction from Research Datasets

Implement a PostgreSQL solution for automated feature extraction and preparation for machine learning models in scientific research. Create a flexible stored procedure that: (1) Dynamically identifies numeric and categorical features, (2) Performs automatic feature scaling and normalization, (3) Generates correlation-based feature selection, and (4) Outputs a machine learning-ready dataset with metadata tracking. Handle datasets with mixed variable types from experimental sciences.