TBDSG Child Abuse/Neglect Recidivism Prediction Results

Community Data Science Results - New ways of getting data-science work done to benefit society. Join TBDSG for presentations of the results of our community-service data-science project.

Backstory
This effort was kicked off at our October 2019 meetup where data scientists interested in collaborating on a community-service project were introduced to community leaders from Gulf-Coast Jewish Family and Community Services (Gulf Coast JFCS) requesting assistance predicting child abuse/neglect recidivism.

What will be presented at this meetup?
Tom Witmer will provide a brief overview of how the problem was constructed and engaged. Those meetup participants delivering solutions will get an opportunity to present lightning talks on their findings and recommendations to Gulf Coast JCFS leadership and the other meetup attendees. Participants should bring whatever visual materials they require to convey their results and insights. An HDMI cable will be available to connect to a display.

Deliverables requested include a predictive model using any technology and recommendations regarding improvements in the information model to support such an analysis.

About the Problem
Children returned to their home after entering Florida’s “system of care” often experience repeated abuse or neglect and must be removed again. An analysis of parental characteristics hypothesized to increase the likelihood of that occurring could assist child-welfare case managers develop plans for wraparound services post-reunification in order to ensure parents continue to grow in their capacity to support their children and avoid additional removals.

What has happened since the kickoff in October 2019?
Eleven people representing individuals as well as companies identified themselves as “Trusted Partners” and received requests for quotation. Solution proposals from these partners were collated to identify methods and technologies proposed. Submissions included: MS Azure, Watson Studio, R, Python, and Tableau among others. Both Microsoft and Amazon assigned support people to be engaged as necessary.

About our Speaker:
Tom Witmer is a data mining instructor at the University of California- Davis School of Continuing and Professional Education and for 10 years worked at Kaiser Permanente where he led the corporate shared services analytics organization supporting 5000 analysts. He is the founder and CEO of SageLegion.