Building a Burnout Early Warning System with Prophet and PyTorch Transformers Using HRV Data

Short summary

Combines Facebook Prophet and PyTorch Transformers to predict burnout from Oura Ring HRV data, shifting from reactive readiness scores to 24-hour-ahead fatigue forecasting. Prophet handles macro trends like weekly workout cycles while a multi-head attention Transformer captures subtle non-linear HRV drops signaling nervous system stress. Includes full code for Oura API data fetching, Prophet baseline forecasting, and a custom Transformer encoder for anomaly detection.

• •Hybrid model pairs Prophet (macro trends) with PyTorch Transformer (micro HRV anomalies) for burnout prediction
• •Uses Oura Ring HRV balance data as the primary autonomic nervous system stress signal
• •Provides complete code for data fetching, baseline forecasting, and Transformer-based sequence learning

Generated with AI, which can make mistakes.