Breath-Dx provides a new benchmark for time series classification in electronic nose technology using the S-OH dataset. It includes 1,234 patients across nine diagnostic groups and specifically accounts for sensor drift in its temporal train/test splits.
HOW THIS AFFECTS YOU
●
researcherYou can now test your models on a large, public dataset that actually accounts for sensor drift.
●
healthThis provides a standardized way to validate portable disease screening tools.