A Mini-Conference to Celebrate Elizabeth Thompson's Contributions to Statistics, Genetics and the University of Washington was held on June 19, 2018.

My talk is titled

Fast Valid Statistical Inference when the Maximum Likelihood Estimate Does Not Exist in an Exponential Family Model and the Usual Asymptotics are Bogus

Reference material for this talk is:

Geyer, C. J. (1990).
Likelihood and Exponential Families.
PhD thesis, University of Washington.
Advisor: Elizabeth A. Thompson

Geyer, C. J. (2009).
Likelihood inference in exponential families and directions of recession.
Electronic Journal of Statistics, 3 259–289.

Eck, D. J. (2017).
Statistical Inference in Multivariate Settings.
PhD thesis, University of Minnesota.

Eck, D. J., and Geyer, C. J. (submitted).
Computationally efficient likelihood inference in exponential families when the maximum likelihood estimator does not exist.
Submitted to Annals of Statistics.

(and, of course, references cited in these works).


The slides for the talk are here: PDF and knitr source.

Supporting Material

Some supporting material that didn't fit on the slides is here: PDF and knitr source.