Project Grant 2515709

This $245,190 Project Grant award from the National Science Foundation's (NSF) Mathematical and Physical Sciences program, with a performance period from July 2025 to June 2028, aims to advance the frontiers of nonparametric Bayesian methodology, theory, and applications. The research program has two overarching objectives: 1) Developing a novel Bayesian inferential framework for generative models using modern machine learning tools to enable statistically principled inference in complex generative systems, and 2) Advancing practical methodology for computerized adaptive testing (CAT) to enhance computer-human interactions through dynamically tailored questioning. Key outcomes include a flexible generative toolkit for simulating conditional distributions, theoretical insights into sparsity-inducing priors, and the integration of machine learning techniques into classical item response theory models for more responsive and individualized assessment tools. This award supports the University of Chicago's expertise in conducting innovative, multidisciplinary research to address critical scientific challenges and national priorities in areas such as artificial intelligence, statistical modeling, and educational assessment.