✅ What is FedOps Fitbit Health Pipeline?

By applying temporal windowing to Fitbit Health device data (including steps, heart rate, calories, and sleep duration) and integrating it with the FedOps federated learning framework, the system enables collaborative training of an LSTM-based sleep quality prediction model across multiple devices while preserving data privacy, thus achieving automated model operations.

Data & Model:

• It automatically reads CSV files (steps, calories, heart rate, sleep), merges them by time and user, calculates a StressLevel feature, and generates fixed-length time windows per user.

• Use the “Fitbit Sleep and Activity Dataset” (https://www.kaggle.com/datasets/arashnic/fitbit) containing:

  heart_rate: Beats per minute (BPM) steps: Daily step count stress_level: Self-reported (1–10)

  sleep_quality: Binary label (0 = poor, 1 = good)