Monitoring and Diagnosis of Depression from Daily Motor Activity

Depression is one of the most common mental disorder associated with suicide. Timely diagnosis and intervention of depression will improve life quality and reduce suicidal rate. Recent studies have shown that motor activities measured from a wearable sensor may correlate with the depression state. I develop a machine learning model to diagnose depression using motor activity data. My model improves the base-line performance by 44%, suggesting the potential of artificial intelligence in mental health management.