Health monitoring system

[AnNguyen87]

Description

This PR implements a comprehensive health monitoring system that combines LSTM autoencoders with machine learning classifiers for behavioral anomaly detection. The system provides multi-level risk assessment (Low/Medium/High) through temporal pattern analysis and real-time behavioral monitoring.

Key Features:
LSTM Autoencoder for detecting anomalies in time-series health data
Dual Classifier System (Random Forest + MLP) for behavioral pattern analysis
Threshold-based alert system for 7 key health metrics (sleep, steps, calories, etc.)
Multi-model risk integration combining temporal and behavioral assessments
Comprehensive CSV outputs including alerts dashboard and top risk cases

Todos

• Tested and working locally
• Code follows the style guidelines of this project
• I have performed a self-review of my code
• Code changes documented
• Requested review from >= 2 devs on the team (one frontend and one backend recommended)

How to test

Run the notebook Alert_System.ipynb in Jupyter/Colab

Run all cells sequentially

Verify no errors occur

Associated MS Planner Tasks

• AI-Powered Behaviour Profile - Model Development (https://teams.microsoft.com/l/entity/com.microsoft.teamspace.tab.planner/planner.v1.1c7aff73-bd5d-4eaf-844f-a0f0914d35b0_p_obnY9eRGNEGXqjBUOhDWKcgAFWt1?tenantId=d02378ec-1688-46d5-8540-1c28b5f470f6&webUrl=https%3A%2F%2Ftasks.teams.microsoft.com%2Fteamsui%2FpersonalApp%2Falltasklists&context=%7B%22subEntityId%22%3A%22%2Fv1%2Fplan%2FobnY9eRGNEGXqjBUOhDWKcgAFWt1%2Fview%2Fboard%2Ftask%2FTGV3uVBU8ke7_23M02BszMgADpJZ%22%2C%22channelId%22%3A%2219%3A02e7479536e247c5a85093cf6dcaca4c%40thread.tacv2%22%7D)

Known Issues

Google Drive Mounting: Requires manual authentication when running in Colab

Large Models: Some model files (.keras, .pkl) are large (>100MB) - may need Git LFS

Memory Intensive: LSTM training requires significant RAM for large datasets

Feature Scaling: Behavioral thresholds may need calibration for different patient demographics