Classifying one-dimensional discrete models with maximum likelihood degree one

TL;DR

This paper classifies all one-dimensional discrete statistical models with maximum likelihood degree one using rational parametrization, introduces fundamental models and chipsplitting games, and shows finiteness results for models in low-dimensional simplices.

Contribution

It provides a complete classification of these models based on their rational parametrizations and introduces a novel combinatorial approach with chipsplitting games.

Findings

All such models can be constructed from fundamental models.

Finitely many fundamental models exist in $ abla_n$ for n ≤ 4.

Chipsplitting games represent fundamental models effectively.

Abstract

We propose a classification of all one-dimensional discrete statistical models with maximum likelihood degree one based on their rational parametrization. We show how all such models can be constructed from members of a smaller class of 'fundamental models' using a finite number of simple operations. We introduce 'chipsplitting games', a class of combinatorial games on a grid which we use to represent fundamental models. This combinatorial perspective enables us to show that there are only finitely many fundamental models in the probability simplex $\Delta_n$ for $n\leq 4$.