Development of Bayesian and frequentist methodology for high-dimensional data, including dimension reduction, regularization, and uncertainty quantification, with a focus on inferential validity in complex modeling regimes.
Design of scalable statistical computing frameworks and algorithms, supported by large-scale simulation and computational experimentation, to enable efficient implementation and rigorous evaluation of modern inferential methods.
Integration of statistical theory with machine learning techniques to develop interpretable, reproducible, and computationally efficient models for data-science applications, including causal inference and network analysis.
Modeling complex life systems: Bayesian inference for Weibull failure times using adaptive MCMC
Statistical Papers, 2026.
Oketch, T., Sepehrifar, M.
A hierarchical Bayesian Weibull model under MNAR censoring with adaptive prior learning for reliable survival inference in small and heterogeneous samples.
Journal of Statistical Planning and Inference (2026 -).
Oketch, T.
Adaptive prior learning in Bayesian frailty and joint survival models with covariates: theoretical and inferential implications.
Journal of Statistical Planning and Inference (2026-).
Oketch, T.
Adaptive Bayesian Weibull survival modeling for high-dimensional multi-omics data with structured shrinkage and biological priors.
Journal of Computational Biology (2025 -).
Oketch, T.
Performance analysis of computational statistics methods for missing value imputation in mass spectrometry-based label-free quantitative proteomics.
BMC Bioinformatics (2025 -).
Oketch, T., Popescu, G.
The rest of my research activities can be found on the links below: