Statistics Seminars 2010-11

We provide a LaTeX template for speakers to use when submitting the abstract for a seminar (sample PDF output here). Using this template makes it straightforward for us to post the details of the talk promptly and without unintended errors that might creep in if we were to transcribe the abstract from another source.

You can receive a weekly announcement of seminar offerings by subscribing to our mailing list at any time! Please follow the link for Statistics Seminars and fill in the fields required to subscribe, unsubscribe, change subscription options, and so on.

Inquiries regarding Seminars should be sent to .

September	October	November	December	January	February	March	April	May	June	July	August

August 2011

Date Speaker Affiliation Title Abstract 08.23 Special Event Emily Fox University of Pennsylvania Bayesian nonparametric methods for complex dynamical phenomena 08.10 Summer Seminar Bob Griffiths University of Oxford Exchangeable pairs of Bernoulli random variables, Krawtchouk polynomials, and Ehrenfest urns 08.09 Special Event Michael Wolf University of Zurich Nonlinear shrinkage estimation of large-dimensional covariance matrices 08.04 Special Thursday Event Akimichi Takemura University of Tokyo Martingales and sequential tests from the viewpoint of game-theoretic probability 08.03 Summer Seminar Persi Diaconis Stanford Statistics & Mathematics Hypergroups, Markov chains, and orthogonal polynomials

July 2011

Date Speaker Affiliation Title Abstract 07.26 Bob Griffiths University of Oxford Importance sampling on evolutionary histories 07.20 Summer Seminar Bob Griffiths University of Oxford Characterization of bivariate distributions and stochastic processes with orthogonal polynomial eigenfunctions, part 3 07.13 Summer Seminar Bob Griffiths University of Oxford Characterization of bivariate distributions and stochastic processes with orthogonal polynomial eigenfunctions, part 2 07.07 Cancelled Eugene Demidenko Dartmouth College Statistical inference for random effects in linear and nonlinear mixed models 07.06 Summer Seminar Bob Griffiths University of Oxford Characterization of bivariate distributions and stochastic processes with orthogonal polynomial eigenfunctions, part 1

June 2011

Date Speaker Affiliation Title Abstract 06.29 Summer Seminar Persi Diaconis Stanford Statistics & Mathematics An actual introduction to orthogonal polynomials 06.28 Tianyi Zhou University of Technology, Sydney GoDec: Randomized low-rank & sparse matrix decomposition in noisy case 06.22 Summer Seminar Persi Diaconis Stanford Statistics & Mathematics Basics of orthogonal polynomials 06.21 Special Event Boaz Nadler Weizmann Institute of Science, Israel On the distribution of Roy's largest root test in MANOVA, signal detection in noise, and canonical correlation analysis 06.07 Special Event Niels Richard Hansen University of Copenhagen Multivariate sparse dynamic process modeling and inference

May 2011

Date Speaker Affiliation Title Abstract 05.31 Howard Wainer National Board of Medical Examiners Uneducated guesses: Three examples of how mistreating missing data yields misguided educational policy 05.24 Yoram Singer Google Entire relaxation path for maximum entropy models 05.17 Noureddine El Karoui UC Berkeley Some robustness questions in high-dimensional statistics 05.10 Robert Gentleman Genentech Three problems in computational biology/statistics 05.03 David Stork Ricoh Labs Efficient, secure, oblivious pattern classification: Classifying patterns that cannot be seen

April 2011

Date Speaker Affiliation Title Abstract 04.26 Ery Arias-Castro UC San Diego Cluster detection in networks using percolation 04.20 Special Wednesday Event Ruslan Salakhutdinov Massachusetts Institute of Technology Learning hierarchical generative models 04.19 Sylvain Sardy University of Geneva Smooth James--Stein model selection for wavelet smoothing, parametric linear model and inverse problem 04.12 SBJC Philip Stark UC Berkeley Close enough for government work: Risk-limiting post-election audits 04.05 Stéphane Mallat École Polytechnique High-dimensional classification with invariant representations

March 2011

Date Speaker Affiliation Title Abstract 03.29 John Ioannidis Stanford Medical School Empirical testing of excess significance bias 03.22 Spring Break Intersession --- No Seminar 03.15 No Seminar 03.09 Special Wednesday Event Mike Baiocchi The Wharton School, University of Pennsylvania Estimating the effect of intensity of care on rates of death for premature infants 03.08 No Seminar 03.03 Special Thursday Event Jian Guo University of Michigan Joint estimation of multiple graphical models 03.02 Special Wednesday Event Xiaochao Xie Harvard University Shrinkage estimation in heteroscedastic hierarchical models 03.01 Susan Holmes Stanford University Statistics Crowd counting a crowd

February 2011

Date Speaker Affiliation Title Abstract 02.24 Special Thursday Event Ryan Adams University of Toronto Tractable inference of probability densities and point processes with Gaussian process priors 02.22 Michael Perlman University of Washington, Seattle Two statistical vignettes: Simpson's paradox and shaved dice 02.16 Special Wednesday Event Imre Risi Kondor California Institute of Technology Non-commutative harmonic analysis in machine learning 02.15 Guenther Walther Stanford University Statistics Inference and modeling with log-concave distributions 02.09 Special Wednesday Event Jelena Bradic Princeton University Regularization for Cox's proportional hazards model with NP-dimensionality 02.08 Larry Brown The Wharton School, University of Pennsylvania Valid statistical inference after model selection 02.02 Special Wednesday Event Allan Sly Microsoft Research, Redmond, WA The reconstruction problem on the tree 02.01 Location Change: Jordan Hall David Donoho Stanford University Statistics Precise optimality in compressed sensing: Rigorous theory and ultra-fast algorithms

January 2011

Date Speaker Affiliation Title Abstract 01.25 Hans Andersen Stanford University  Chemistry The elastic net method in computational statistical mechanics 01.18 Eric Xing Carnegie Mellon University Learning varying coefficient varying structure models: Reverse engineering rewiring networks underlying dynamics processes 01.11 Robert Tibshirani Stanford University  Statistics Strong rules for discarding predictors in lasso-type problems 01.04 Dimitris Politis UC San Diego Model-free model-fitting and predictive distributions

December 2010

Date Speaker Affiliation Title Abstract 12.07 Bradley Efron Stanford University Statistics Tweedie's formula and selection bias

November 2010

Date Speaker Affiliation Title Abstract 11.30 Mark Low University of Pennsylvania Estimating the L1 norm of a multivariate mean vector 11.23 John Chambers Stanford University Statistics R, other languages and object-oriented programming 11.15 Monday Judea Pearl UCLA Causal inference: Mathematical foundations and practical applications 11.09 Gabor Szekely National Science Foundation Testing independence for high-dimensional data with application to time series 11.02 BSJC Persi Diaconis Stanford University Statistics Estimating n parameters based on a sample of size one

October 2010

Date Speaker Affiliation Title Abstract 10.26 Xiao-Li Meng Harvard University A seminal trio: Applying Efron's (1967) self-consistency principle to Donoho and Johnstone's (1994) wavelets and Tibshirani's (1996) LASSO regressions with incomplete and irregularly spaced data 10.19 Wing Wong Stanford University Statistics Long range haplotyping and its applications 10.12 Abraham Wyner University of Pennsylvania A statistical analysis of multiple temperature proxies: Are reconstructions of surface temperatures over the last 1000 years reliable? 10.05 Various Special Seminar Series in Honor of Richard Olshen

September 2010

Date Speaker Affiliation Title Abstract 09.28 Stephen Boyd Stanford University Electrical Engineering Distributed optimization and statistical learning via the alternating direction method of multipliers 09.21 John Ionnadis Stanford University School of Medicine Network meta-analysis and meta-analysis networks

SBJC = Stanford/Berkeley Joint Colloquium at Stanford

BSJC = Berkeley/Stanford Joint Colloquium at Berkeley