Decoding the H-likelihood

TL;DR

This paper discusses the concept of H-likelihood, a statistical approach for models with unobservable variables, providing insights into likelihood inference methods.

Contribution

It offers a new perspective on likelihood inference for models with unobservables through the framework of H-likelihood, expanding on prior work by Lee and Nelder.

Findings

Clarifies the theoretical foundations of H-likelihood

Demonstrates advantages over traditional likelihood methods

Provides examples illustrating the application of H-likelihood

Abstract

Discussion of "Likelihood Inference for Models with Unobservables: Another View" by Youngjo Lee and John A. Nelder [arXiv:1010.0303]