School of Applied Data
Associate Professor of Applied Data
About
Okonkwo designs evaluation methods that still work when civic datasets are sparse, messy, or politically sensitive. Before Cedarhaven she spent four years at a municipal analytics office, where she learned that a model nobody can explain will not survive a public-comment period. Her studio treats documentation—codebooks, model cards, partner memos—as part of the science, not an afterthought.
Education
- Ph.D., Machine Learning, Carnegie Mellon University
- M.S., Statistics, University of Michigan
- B.S., Mathematics, Howard University
Courses
Research
Grouped cross-validation for administrative data; uncertainty language for non-technical boards; ethics of predictive tools in housing and public health.