Adrian E. Raftery, University of Washington

Statistical Inference with Model Uncertainty

Keynote Speaker

Location: Vencovského aula (New Building) 

Choosing a statistical model and accounting for uncertainty about this choice are important parts of the scientific process and are required for common statistical tasks such as parameter estimation, interval estimation, statistical inference, point prediction, and interval prediction. A canonical example is the choice of variables in a linear regression model. Many ways of doing this have been proposed, including Bayesian and penalized regression methods, and it is not clear which are best. We compare 21 popular methods via an extensive simulation study based on a wide range of real datasets. We found that three adaptive Bayesian model averaging methods performed best across all the statistical tasks and that two of these were also among the most computationally efficient. We also compared different priors on model space. Finally we addressed the question of whether model averaging provides an advantage over model selection. This is joint work with Anupreet Porwal.

About the Speaker

Adrian E. Raftery is the Boeing International Professor of Statistics and Sociology, and an adjunct professor of Atmospheric Sciences, at the University of Washington in Seattle. He was born in Dublin, Ireland, and obtained a B.A. in Mathematics (1976) and an M.Sc. in Statistics and Operations Research (1977) at Trinity College Dublin. He obtained a doctorate in mathematical statistics in 1980 from the Université Pierre et Marie Curie in Paris, France under the supervision of Paul Deheuvels. He was a lecturer in statistics at Trinity College Dublin from 1980 to 1986, and then an associate (1986-1990) and full (1990-present) professor of statistics and sociology at the University of Washington. He was the founding Director of the Center for Statistics and Social Sciences (1999-2009).

Raftery has published over 200 articles in peer-reviewed statistical, sociological and other journals. His research focuses on Bayesian model selection and Bayesian model averaging, model-based clustering, inference for deterministic simulation models, and the development of new statistical methods for demography, sociology, and the environmental and health sciences.

He is a member of the United States National Academy of Sciences, a Fellow of the American Academy of Arts and Sciences, an Honorary Member of the Royal Irish Academy, a member of the Washington State Academy of Sciences, a Fellow of the American Statistical Association, a Fellow of the Institute of Mathematical Statistics, and an elected Member of the Sociological Research Association. He has won the Population Association of America’s Clifford C. Clogg Award, the American Sociological Association’s Paul F. Lazarsfeld Award for Distinguished Contribution to Knowledge, the Jerome Sacks Award for Outstanding Cross-Disciplinary Research from the National Institute of Statistical Sciences, the Parzen Prize for Statistical Innovation, and the Science Foundation Ireland St. Patrick’s Day Medal. He is also a former Coordinating and Applications Editor of the Journal of the American Statistical Association and a former Editor of Sociological Methodology. He was identified as the world’s most cited researcher in mathematics for the decade 1995-2005 by Thomson-ISI.

Thirty-two students have obtained Ph.D.’s working under Raftery’s supervision, of whom 21 hold or have held tenure-track university faculty positions. He has a total of 128 academic descendants.

Log in