- Research Name Student Yield & Success
- Client Wayne State University
- Category PRACTICUM PROJECTS
- RESEARCH YEAR 2019
The goal of this project is to tap into the power of advanced data analytics and to develop algorithms for forecasting/predictive engine for the WSU enrollment management offices of undergraduate admissions and student financial aid.
Predictive models were built to help with the two metrics, Yield and Persistence. These models would be linked to WSU’s in-house business intelligence tool, Power BI, providing admission counselors with information on expected yield by high school, zip code, territory, etc. Through the use of Shapley Additive explanations (SHAP) predictions, the team was able to provide the important metrics for each feature and customized information for each applicant as to why they are likely or unlikely to Yield or Persist (depending on the model being used).
Abhinav Reddy Nalla, Afsana Kabir, Benjamin Messick, Praveen Kumar, and Xiaoxu Zhang