A few weeks ago I released a synthetic sleep health dataset on Kaggle and it took off faster than I expected. Sharing it here in case anyone finds it useful.<\/p>\n
What’s in it:<\/p>\n
– 100,000 records, 32 features, 3 prediction targets<\/p>\n
– Sleep architecture: REM %, deep sleep %, latency, wake episodes<\/p>\n
– Lifestyle: caffeine, alcohol, screen time, exercise, steps<\/p>\n
– Psychological: stress score, chronotype, mental health condition<\/p>\n
– Demographics: 12 occupations, 15 countries, ages 18-69<\/p>\n
Three ML targets:<\/p>\n
– cognitive_performance_score- regression (0\u2013100)<\/p>\n
– sleep_disorder_risk – multiclass (Healthy \/ Mild \/ Moderate \/ Severe)<\/p>\n
– felt_rested – binary classification<\/p>\n
One finding that surprised people:<\/p>\n
Lawyers average 5.74 hrs of sleep and 7.3\/10 stress. Retired individuals average 8.03 hrs and 2.6\/10 stress. That 2.13-hour gap shows up clearly in every model – occupation is the strongest predictor of sleep health in the entire dataset.<\/p>\n
All distributions are calibrated against CDC, Sleep Foundation, and Frontiers in Sleep research. Correlations match peer-reviewed values (e.g. stress vs quality r=-0.64).<\/p>\n
Link in profile if you want to check it out. Happy to answer questions about how it was built.<\/p>\n<\/div>\n
submitted by \/u\/Mohan137 <\/a>
[link]<\/a><\/span> [comments]<\/a><\/span><\/p>
