|
Activity Number:
|
436
- Gottfried E. Noether Lectures
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 6, 2020 : 10:00 AM to 11:50 AM
|
|
Sponsor:
|
Noether Award Committee
|
|
Abstract #314421
|
|
|
Title:
|
Empirical Likelihood: Some History and New Directions
|
|
Author(s):
|
Art Owen*
|
|
Companies:
|
Stanford University
|
|
Keywords:
|
|
|
Abstract:
|
The likelihood function runs all through statistics with a central role in estimation (the MLE) and testing (Wilks' theorem) and strong optimality properties. The likelihood function usually makes strong assumptions that the data come from a known parametric distribution with only some finite dimensional parameter unknown. The 1940s and 1950s saw a surge of interest in nonparametric methods, with Gottfried Noether being one of the leaders. Some nonparametric MLEs were developed. The empirical likelihood is primarily a nonparametric likelihood ratio method. It satisfies a version of Wilks' theorem that we can use without needing a parametric model. This talk will present the basic theory and some use cases and point to recent developments, especially Bayesian empirical likelihood.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2020 program
|
