in brief
We conduct cutting-edge research developing Bayesian statistical and machine learning methods for the analysis of various types of data including genomic, electronic health, and public health data.
Our lab is founded on collaborations with both accademia and industry, please contact us if you are interested in potential collaboration.
long(er) form
If it has cool math and an impactful question, we are interested. Based on Dr. Silverman’s combined medical and statistical training, we tend to gravitate to problems in the analysis of biomedical data; especially, genomics and high-throughput assays. However, our research interests are varied and include both theoretical and applied aspects of mathematics and statistics.
From methodological perspective, we have particular interest in developing robust inferential and predictive models for non-identifiable problems. This includes issues of unmeasured confounding, systematic measurement bias, model misspecification, and scale reliant inference.
example dissertation projects
- Bayesian partially identified models lead to dramatic decreases in Type-I and II errors in microbiome and gene expression analyses. link
- Syndromic surveillance reveals widespread underestimating in COVID-19 prevalence within the US in 2020. link
- Scalable Bayesian multinomial logistic normal models enhance analyses of microbiome data. link
- Bayesian conformal prediction and out-of-sample boosting leads to more efficient prediction intervals and improved conditional coverage. (in progress)
- Conditional generative models for analyzing microbiome data directly in sequence space. (in progress)