Title: An attempt to predict suicides using smartphones
Summary: Psychologists at Harvard’s psychology department have been trying to use sensors and a tracker app to collect data, ultimately used to predict who is likely to attempt suicide and when to intervene. In theory, with enough data, a good machine learning model can discern signals that psychiatrists normally can’t see. This, however, is not feasible in the field of suicide intervention since suicide is such a rare event, even among those at the highest risk. Therefore, any prediction can result in a false positive, which pushes legal responsibilities onto clinicians. There are also conversations about the privacy of patients whose lives are peered into. On the bright side, these prediction algorithms have proven to be more accurate than traditional methods. Furthermore, it is stated that Ms. Cruz, one of the volunteers in this study, thinks it is almost easier to tell the truth to a computer. Overall, it is clear that machine learning applications can help us prevent suicide attempts, but they should not perfectly substitute for a clinician’s care.