|
Abstract:
|
Meeting the needs of people experiencing multi-faceted difficulties (such as homelessness, chronic medical conditions, and mental illness) requires coordinated care across systems. This paper describes our work building a prototype system for Johnson County, Kansas that uses machine learning to prioritize outreach to individuals with a history of mental illness most at risk of being booked into jail within the next year, by bringing together jail, emergency medical, and mental health data. Our system out-performed both a random baseline and simple heuristics and is one that domain experts are likely to use and implement.
|
