Valid and efficient imprecise-probabilistic inference with partial priors, III. Marginalization

TL;DR

This paper develops a new inferential model framework for imprecise-probabilistic inference that effectively handles nuisance parameters through marginalization, offering valid and efficient solutions even in complex models.

Contribution

It introduces a profile-likelihood-based marginal IM approach that guarantees validity and efficiency in both parametric and non-parametric inference, connecting to conformal prediction.

Findings

Profile likelihood is central to marginal inference with nuisance parameters.

The proposed IM approach is valid and efficient even when likelihoods don't factor nicely.

Applications include Behrens--Fisher, gamma mean problems, and non-parametric risk minimizers.

Abstract

As Basu (1977) writes, "Eliminating nuisance parameters from a model is universally recognized as a major problem of statistics," but after more than 50 years since Basu wrote these words, the two mainstream schools of thought in statistics have yet to solve the problem. Fortunately, the two mainstream frameworks aren't the only options. This series of papers rigorously develops a new and very general inferential model (IM) framework for imprecise-probabilistic statistical inference that is provably valid and efficient, while simultaneously accommodating incomplete or partial prior information about the relevant unknowns when it's available. The present paper, Part III in the series, tackles the marginal inference problem. Part II showed that, for parametric models, the likelihood function naturally plays a central role and, here, when nuisance parameters are present, the same principles suggest that the profile likelihood is the key player. When the likelihood factors nicely, so that the interest and nuisance parameters are perfectly separated, the valid and efficient profile-based marginal IM solution is immediate. But even when the likelihood doesn't factor nicely, the same profile-based solution remains valid and leads to efficiency gains. This is demonstrated in several examples, including the famous Behrens--Fisher and gamma mean problems, where I claim the proposed IM solution is the best solution available. Remarkably, the same profiling-based construction offers validity guarantees in the prediction and non-parametric inference problems. Finally, I show how a broader view of this new IM construction can handle non-parametric inference on risk minimizers and makes a connection between non-parametric IMs and conformal prediction.