Statistical Inference for Markov Chains with Known Structure

In this talk, we will discuss nonparametric estimation for Markov chains that satisfy stochastic recursions of a known form, but in which the driving noise has an unknown distribution. We assume that we observe samples of the driving noise. We will argue that such Markov chains/data sets arise naturally within the setting of operations research and management science. We will discuss general statistical theory for such Markov chains, including consistency of “plug-in” estimators, the influence function, central limit theorems, and related computational issues. This work is joint with Lin Fan.

Bio: Professor Peter Glynn is a Thomas W. Ford Professor in the School of Engineering, Professor in the Department of Management Science and Engineering and the Institute for Computational and Mathematical Engineering, and Professor (by courtesy) in the Department of Electrical Engineering, all at Stanford University.

Share this Event