Ruixun Zhang
Peking University
Room: M3 3127
Benjamin Landon
University of Toronto
Room: M3 3127
Distinguished Lecture Series
Paul Gustafson
Department of Statistics at the University of British Columbia
Room: DC 1302
Bayesian Inference when Parameter Identification is Lacking: A Narrative Arc across Applications, Methods, and Theory
Partially identified models generally yield “in between” statistical behavior. As the sample size goes to infinity, the posterior distribution on the target parameter heads to a distribution narrower than the prior distribution but wider than a single point. Such models arise naturally in many areas, including the health sciences. They arise particularly when we own up to limitations in how data are acquired. I aim to highlight the narrative arc associated with partial identification. This runs from the applied (e.g., broaching the topic with subject-area scientists), to the methodological (e.g., implementing a Bayesian analysis without full identification), to the theoretical (e.g., characterizing what is going on as generally as possible). As per many areas of statistics, there is good scope to get involved across the whole arc, rather than just at one end or other.
Babak Shahbaba
University of California, Irvine (UCI)
Room: M3 3127
Luhao Zhang
Columbia University
Room: M3 3127
Actuarial Science and Financial Mathematics seminar seriesJuliana Schulz Room: M3 3127 |
Actuarial Science and Financial Mathematics seminar seriesAnastasis Kratsios Room: M3 3127 |
Christopher Chambers
Georgetown University
Room: M3 3127
Distinguished Lecture Series
Bhramar Mukherjee
John D Kalbfleisch Distinguished University Professor of Biostatistics
Siobán D. Harlow Collegiate Professor of Public Health
Chair of the Department of Biostatistics
University of Michigan
Room: DC 1302
The Data Struggle of the Unseen
Despite several proposed roadmaps to increase diversity in scientific research, most of the world's research data are collected on people of European ancestry. We rely on summary statistics from historically privileged populations and then devise clever statistical methods to transfer/transport them for cross-ancestry use. In this talk, I would first argue the obvious: for building fair algorithms we need fair training datasets. However, till we have reached the dream of equitable big data at a global scale, statisticians have an important role to play. In fact we have the perfect tools to study the "unobserved" through modeling of missing data, selection bias and alike. I will share examples from my personal journey as a statistician where doing good and timely statistical work with imperfect data quantified important disparity in health outcomes and led to policy impact. I will conclude the talk with a call to arms for statisticians to lead efforts for creating, curating, collecting data and pioneering new scientific studies, not just remain on the design and analytic fringes. As public health statisticians, our job is not just to predict, but to prevent. The talk is based on years of work with my students and colleagues at the Department of Biostatistics, University of Michigan and inspired by the transformative experience we shared as a statistical team working on the COVID-19 pandemic.
Donglin Zeng
University of Michigan
Room: M3 3127