Display Accessibility Tools

Accessibility Tools

Grayscale

Highlight Links

Change Contrast

Increase Text Size

Increase Letter Spacing

Dyslexia Friendly Font

Increase Cursor Size

Colloquium - Veronika Rockova

Institution: University of Chicago
Title: Metropolis-Hastings via Classification
Date: February 1, 2022
Location: Zoom (Click here for meeting details)
Time: 10:20 AM - 11:10 AM Eastern Time

Abstract:
This paper develops a Bayesian computational platform at the interface between posterior sampling and optimization in models whose marginal likelihoods are difficult to evaluate. Inspired by contrastive learning and Generative Adversarial Networks (GAN), we reframe the likelihood function estimation problem as a classification problem. Pitting a Generator, who simulates fake data, against a Classifier, who tries to distinguish them from the real data, one obtains likelihood (ratio) estimators which can be plugged into the Metropolis-Hastings algorithm. The resulting Markov chains generate, at a steady state, samples from an approximate posterior whose asymptotic properties we characterize. Drawing upon connections with empirical Bayes and Bayesian mis-specification, we quantify the convergence rate in terms of the contraction speed of the actual posterior and the convergence rate of the Classifier. Asymptotic normality results are also provided which justify the inferential potential of our approach. We illustrate the usefulness of our approach on examples which have challenged for existing Bayesian likelihood-free approaches.

Click here to view the colloquium flyer

news

May 3, 2022 Yuehua Cui elected as a Fellow of American Statistical Association

Yuehua Cui, Professor and Director of Graduate Programs in the Department of Statistics and Probability (STT), has been named an elected Fellow of the American Statistical Association (ASA). 

April 19, 2022 Dr. Huan Lei honored with NSF Early CAREER Award

Huan Lei, assistant professor in the Department of Statistics and Probability (STT) is among five Michigan State University (MSU) researchers from the College of Natural Science (NatSci) who have received NSF Early CAREER Faculty Awards. Their cutting-edge research pushes the limits of science, while their devotion to education is preparing the next generation of scientists to propel their field even further. Collectively, over the next five years (2022-2027) they will receive more than $3.7 million in National Science Foundation (NSF) funding.

March 18, 2022 STT PhD student awarded IMPACTS fellowship for 2021-2022

Sikta Das Adhikari, in her third year with the STT PhD program, was awarded an IMPACTS fellowship for the 2021-2022 academic year. Integrated training Model in Plant And Compu-Tational Sciences (IMPACTS) is an NSF-funded program for training doctoral students to employ advanced computational/data science approaches to address grand challenges in plant biology.

March 17, 2022 Dr. Maiti named AAAS fellow for his contributions to statistics and data science

Tapabrata “Taps” Maiti, MSU Foundation Professor in the Department of Statistics and Probability, was recently named an American Association for the Advancement of Science, or AAAS, fellow for his distinguished contributions to the fields of statistics and data science, particularly for contributions to data-driven discovery, and for outstanding teaching and training of the next generation of data scientists.

upcoming events