Dynamic Segmentation Engine with Machine Learning

Build an advanced customer segmentation system that dynamically adapts clustering algorithms based on data characteristics. The solution should automatically select and tune clustering algorithms (K-means, DBSCAN, Gaussian Mixture) using silhouette scoring and information-theoretic criteria. Implement a feature engineering pipeline that can handle mixed data types, perform automated feature selection, and generate interpretable segment profiles. Include mechanisms for tracking segment stability and detecting emerging customer micro-segments.