Saved in:
Bibliographic Details
Main Author: Wang, Adam
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2404.02167
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913295907160064
author Wang, Adam
author_facet Wang, Adam
contents The following note proves that conditional entropy of a sequence is almost time-reversal invariant, specifically they only differ by a small constant factor dependent only upon the forward and backward models that the entropies are being calculated with respect to. This gives rise to a numerical value that quantifies learnability, as well as a methodology to control for distributional shift between datasets. Rough guidelines are given for practitioners.
format Preprint
id arxiv_https___arxiv_org_abs_2404_02167
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A remark on conditional entropy
Wang, Adam
Information Theory
The following note proves that conditional entropy of a sequence is almost time-reversal invariant, specifically they only differ by a small constant factor dependent only upon the forward and backward models that the entropies are being calculated with respect to. This gives rise to a numerical value that quantifies learnability, as well as a methodology to control for distributional shift between datasets. Rough guidelines are given for practitioners.
title A remark on conditional entropy
topic Information Theory
url https://arxiv.org/abs/2404.02167