Spectral-Analyze Wearable Sleep Data for a Healthtech Pilot
Overview
What this challenge is about.
You receive 30 nights of wearable data per 25 volunteers, with polysomnography-derived ground-truth stages (Wake / NREM / REM). Engineer spectral features (delta, theta, alpha, beta band power; spectral entropy; wavelet coefficients), train a small classifier (random forest or logistic regression) on the existing morphological features vs. the new spectral set, and quantify lift in macro-F1. Plot per-feature importance and provide medical-advisory-board-friendly explanations. Deliver the notebook + 3-page methodology write-up.
The Brief
What you'll do, and what you'll demonstrate.
Quantify the macro-F1 lift from spectral features for sleep-stage classification, with each new feature explainable to a non-engineer medical board.
Earning criteria — what you'll demonstrate
- Engineer spectral and wavelet features from physiological time series
- Quantify feature-group contribution rigorously (not via single-feature ablation alone)
- Communicate technical features to a non-engineer medical audience
- Document methodology for medical-advisory-board review
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
Careers
Roles this prepares you for.
Real titles. Real skill bridges. Pick the one closest to your trajectory.
Data Scientist
Spectral feature engineering with rigorous lift measurement and medical-board-friendly explanations is exactly the kind of work data scientists ship in consumer healthtech.
This challenge sharpens
- spectral-analysis
- feature-engineering
- data-storytelling
Applied AI Scientist
Translating signal-processing methods into explainable features for a medical advisory board is the applied-AI scientist's daily craft.
This challenge sharpens
- spectral-analysis
- wavelet-analysis
- classification
Research Scientist
Documenting a methodology write-up for board-grade review is part of every junior research scientist's first year in health-AI.
This challenge sharpens
- spectral-analysis
- feature-engineering
- wavelet-analysis