Hypothesis testing with active information

TL;DR

This paper introduces a novel hypothesis testing framework based on active information, providing exact probabilities of type-I errors, which is a first in this research area.

Contribution

It develops the first method to derive exact type-I error probabilities for hypothesis tests involving active information.

Findings

Exact type-I error probabilities are derived for the first time.

The approach advances hypothesis testing in information theory.

Provides a new statistical tool for analyzing active information.

Abstract

We develop hypothesis testing for active information -the averaged quantity in the Kullback-Liebler divergence. To our knowledge, this is the first paper to derive exact probabilities of type-I errors for hypothesis testing in the area.