On the Behrens--Fisher problem: A globally convergent algorithm and a finite-sample study of the Wald, LR and LM Tests

TL;DR

This paper introduces a globally convergent algorithm for the multivariate Gaussian Behrens--Fisher problem, enabling finite-sample analysis of classical tests and extending applicability to high-dimensional data.

Contribution

It provides a provably convergent algorithm for the Behrens--Fisher problem and explores finite-sample properties of Wald, LR, and LM tests in high dimensions.

Findings

Algorithm achieves global convergence and handles high-dimensional data.

Finite-sample size and power of tests are characterized beyond asymptotic results.

Systematic algebraic relation between Wald, LR, and LM tests established.

Abstract

In this paper we provide a provably convergent algorithm for the multivariate Gaussian Maximum Likelihood version of the Behrens--Fisher Problem. Our work builds upon a formulation of the log-likelihood function proposed by Buot and Richards \citeBR. Instead of focusing on the first order optimality conditions, the algorithm aims directly for the maximization of the log-likelihood function itself to achieve a global solution. Convergence proof and complexity estimates are provided for the algorithm. Computational experiments illustrate the applicability of such methods to high-dimensional data. We also discuss how to extend the proposed methodology to a broader class of problems. We establish a systematic algebraic relation between the Wald, Likelihood Ratio and Lagrangian Multiplier Test ($W\geq \mathit{LR}\geq \mathit{LM}$) in the context of the Behrens--Fisher Problem. Moreover, we use our algorithm to computationally investigate the finite-sample size and power of the Wald, Likelihood Ratio and Lagrange Multiplier Tests, which previously were only available through asymptotic results. The methods developed here are applicable to much higher dimensional settings than the ones available in the literature. This allows us to better capture the role of high dimensionality on the actual size and power of the tests for finite samples.