Machine Learning Anomaly Detection in Sensor Streams

Develop a real-time anomaly detection framework for scientific sensor data using scikit-learn and Dask for distributed computing. The system must: 1) Support multiple sensor input streams with variable sampling rates, 2) Implement adaptive Isolation Forest and Local Outlier Factor algorithms, 3) Generate configurable alert thresholds, and 4) Create a Flask-based dashboard for monitoring and visualization. Include robust error handling and support for streaming data from environmental monitoring sensors.