Wearable Health Data Anomaly Detection System

Build a robust Python system using time-series analysis techniques and scikit-learn that can detect subtle health anomalies from continuous wearable device data streams. The solution must handle multivariate sensor inputs, implement advanced statistical and machine learning-based anomaly detection algorithms, and generate contextual health risk alerts. Support multiple wearable device protocols, handle data from diverse sensor types, and provide a flexible architecture for integrating new detection models.