Regret and Jeffreys Integrals in Exp. Families

TL;DR

This paper investigates the conditions under which minimax redundancy, minimax regret, and Jeffreys integrals are finite or infinite within exponential families, addressing fundamental questions in statistical theory.

Contribution

It provides a detailed analysis of the finiteness conditions for minimax redundancy, regret, and Jeffreys integrals in exponential families, clarifying their interrelations.

Findings

Identifies conditions for finiteness of Jeffreys integrals.

Establishes links between minimax regret and Jeffreys integrals.

Provides criteria for when redundancy measures are finite.

Abstract

The problem of whether minimax redundancy, minimax regret and Jeffreys integrals are finite or infinite are discussed.