Professional Summary
Fifth-year Ph.D. candidate in Statistics with a focus on high-dimensional and Bayesian modeling. My research centers on developing scalable and robust variable selection methods for generalized linear models, particularly in biomedical and public health applications. I am especially interested in building interpretable models for complex, high-dimensional datasets, integrating Bayesian sparsity learning and a joint modeling framework. My dissertation focuses on developing computationally efficient and robust methods for variable selection in high-dimensional binary regression, with attention to challenges like model misspecification and missing data.
Research Support
08-2025 -04-2026 Taft Research Center - University of Cincinnati Taft Dissertation Fellowship Active Type:Fellowship Level:University
Publications
Electronic Journal
Eric Odoom and Xia Wang (2026) Consistent and scalable variable selection with robust link functions .New England Journal of Statistics & Data Science,
Eric Odoom, Ouyang Jiarong Cao Xua & Wang Xia (2026) Hierarchical skinny Gibbs sampler in logistic regression using Pólya-Gamma latent variables .Statistics and Its Interface, 19 (2 ) ,
Presentations
Invited Presentations
Eric Odoom and Xia Wang (05-2026. ) (In Press. ) Consistent and scalable variable selection with robust link functions .University of Connecticut, Storrs, CT. Conference. . Level:Regional
Poster Presentations
Eric Odoom (08-2024. ) (Under Review. ) Bayesian Dirichlet regression for correlated compositional outcomes with application to an experimental sleep study .Oregon Convention Center, Portland, OR. . Conference. . Level:National
Paper Presentations
Eric Odoom (08-2025. ) Hierarchical skinny Gibbs sampler in logistic regression using Pólya -gamma latent variables .Nashville, TN. Conference. Level:National
Eric Odoom (03-2025. ) Hierarchical skinny Gibbs sampler in logistic regression using Pólya -gamma latent variables .New Orleans, LA. Conference. Level:National