Kliakhandler Lecture Series (10/16) - Richard Davis, Columbia University

The Department of Mathematical Sciences is delighted to welcome Dr. Richard Davis from Columbia University to deliver the ninth annual Kliakhandler Lecture Series, Adventures in Statistics. The Series will be held October 15 - 16, 2026.

Research Seminar:
Friday, October 16, 2026
1:00 p.m.
Fisher 101

Title: The Magic of Sample Splitting in Time Series Model Checking

Abstract: Whenever we fit a statistical model to data that unfolds over time — stock prices, weather, sensor readings — we want to know if the model actually did its job. One standard way to check is to look at the "leftovers": whatever pattern is left over after the model has explained what it can. If the model is good, those leftovers should look like pure randomness, with no structure left to find.

The trouble is, the very act of fitting the model contaminates the leftovers — they end up secretly correlated with each other in a way that makes them look "too random" or "not random enough," even when the model is perfectly correct. This has forced statisticians to build complicated corrections just to interpret what they're seeing.

This talk describes a much simpler fix: use the first half of the data to fit the model. Then use the entire data set to compute residuals. Then the contamination problem disappears entirely, and the leftovers behave exactly as if they were pure, untouched randomness. We'll build intuition for why this trick works, see it play out in simulations, and show that it leads to model-checking methods that are simpler to use and often more effective at catching a bad model than standard approaches.

​​​​​​​Biography: Richard Davis is the Howard Levene Professor of Statistics at Columbia University and former chair of the Statistics Department (2013-19). He has held academic positions at MIT, Colorado State University, and visiting appointments at numerous other universities. He was Hans Fischer Senior Fellow at the Technical University of Munich (2009-12) , Villum Kan Rasmussen Visiting Professor (2011-13) at the University of Copenhagen , and Jubilee Professor at Chalmers University (2019). Davis is a fellow of the Institute of Mathematical Statistics, the American Statistical Association, and is an elected member of the International Statistical Institute. He was president of IMS in 2016 and Editor-in-Chief of Bernoulli Journal 2010-12. He is co-author (with Peter Brockwell) of the best-selling books, "Time Series: Theory and Methods", "Introduction to Time Series and Forecasting", and the time series analysis computer software package, ITSM2000. Together with Torben Andersen, Jens-Peter Kreiss, and Thomas Mikosch, he co-edited the "Handbook in Financial Time Series" and with Holan, Lund, and Ravishanker, the book, "Handbook of Discrete-Valued Time Series". In 1998, he won (with collaborator W.T.M Dunsmuir) the Koopmans Prize for Econometric Theory. He has advised/co-advised 36 PhD students. His research interests include time series, applied probability, extreme value theory, spatial-temporal modeling, and developing ML tools in extreme value theory. More information about Professor Davis is available at https://sites.stat.columbia.edu/rdavis/.