A new AI model trained on over 2.5 billion hours of Apple Watch data may predict health conditions better than traditional methods.
A new research paper has been published on an AI model trained on Apple Watch data, which can now predict health conditions more accurately. Apparently, it’s more accurate than traditional sensor-based approaches, according to a newly published study.
The research paper is titled „Beyond Sensor Data: Foundation Models of Behavioral Data from Wearables Improve Health Predictions“. The study shows a machine learning model that analyzes user behavior with the aim of flagging potential health issues.
Earlier models of this kind were focused on sensor outputs (in real time) like heart rate or blood oxygen. However, this new model also identifies patterns in how you move, sleep, or exercise over time.
