Effects of Sequence Partitioning on Compression Rate

TL;DR

This paper investigates how dividing data sequences into parts affects compression efficiency, proving that optimal partitioning can reduce entropy rates and improve overall compression performance.

Contribution

It introduces a theoretical framework for sequence partitioning, demonstrating how to find partitions that lower entropy rates and optimize compression.

Findings

Partitioning can reduce the entropy rate of subsequences.

An optimization problem for optimal partitioning is formulated.

Partitioning improves overall compression rate.

Abstract

In the paper, a theoretical work is done for investigating effects of splitting data sequence into packs of data set. We proved that a partitioning of data sequence is possible to find such that the entropy rate at each subsequence is lower than entropy rate of the source. Effects of sequence partitioning on overall compression rate are argued on the bases of partitioning statistics, and then, an optimization problem for an optimal partition is defined to improve overall compression rate of a sequence.