1 The Problem
We took on Design Challenge #1 to analyze the problem of identifying optimal foster families using
relevant literature to guide our approach.
2 Our Assumptions
We defined a successful placement to be a discharge reason of any of the four possible permanency outcomes. This limited us to a sample size of 6,008 cases in Manhattan from 2015_AFCARS_Foster_Data, 1,454 of which resulted in a permanency outcome.
Our chosen model parameters are also based on the most important family and child characteristics in literature.
3 Our Approach
After exploring, cleaning, and clustering the data into intuitive groups, we created a logistic regression model that could predict the probability of a permanency outcome.