Generalized Estimating Equation for the Student-t Distributions

TL;DR

This paper extends the generalized maximum likelihood estimation framework to the Student-t distribution within the $ ext{M}^{( ext{alpha})}$-family, providing new estimators and broadening the applicability of relative $ ext{alpha}$-entropy methods.

Contribution

It generalizes the $ ext{M}^{( ext{alpha})}$-family to include multivariate and continuous distributions like Student-t, and extends estimation techniques accordingly.

Findings

Student-t distributions are shown to belong to the generalized $ ext{M}^{( ext{alpha})}$-family.

The paper derives new generalized estimators for Student-t parameters.

Extension of orthogonality and estimation results to broader distribution families.

Abstract

In \cite{KumarS15J2}, it was shown that a generalized maximum likelihood estimation problem on a (canonical) $\alpha$-power-law model ($\mathbb{M}^{(\alpha)}$-family) can be solved by solving a system of linear equations. This was due to an orthogonality relationship between the $\mathbb{M}^{(\alpha)}$-family and a linear family with respect to the relative $\alpha$-entropy (or the $\mathscr{I}_\alpha$-divergence). Relative $\alpha$-entropy is a generalization of the usual relative entropy (or the Kullback-Leibler divergence). $\mathbb{M}^{(\alpha)}$-family is a generalization of the usual exponential family. In this paper, we first generalize the $\mathbb{M}^{(\alpha)}$-family including the multivariate, continuous case and show that the Student-t distributions fall in this family. We then extend the above stated result of \cite{KumarS15J2} to the general $\mathbb{M}^{(\alpha)}$-family. Finally we apply this result to the Student-t distribution and find generalized estimators for its parameters.