Neural Markovian predictive compression: An algorithm for online lossless data compression

Erez Shermer, Mireille Avigal, Dana Shapira

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Scopus citations

Abstract

This work proposes a novel practical and general-purpose lossless compression algorithm named Neural Markovian Predictive Compression (NMPC), based on a novel com-bination of Bayesian Neural Networks (BNNs) and Hidden Markov Models (HMM). The result is an interesting combination of properties: Linear processing time, constant memory storage performance and great adaptability to parallelism. Though not limited for such uses, when used for online compression (compressing streaming inputs without the latency of collecting blocks) it often produces superior results compared to other algorithms for this purpose. It is also a natural algorithm to be implemented on parallel platforms such as FPGA chips.

Original languageEnglish
Title of host publicationProceedings - Data Compression Conference, DCC 2010
Pages209-218
Number of pages10
DOIs
StatePublished - 2010
Externally publishedYes
EventData Compression Conference, DCC 2010 - Snowbird, UT, United States
Duration: 24 Mar 201026 Mar 2010

Publication series

NameData Compression Conference Proceedings
ISSN (Print)1068-0314

Conference

ConferenceData Compression Conference, DCC 2010
Country/TerritoryUnited States
CitySnowbird, UT
Period24/03/1026/03/10

Fingerprint

Dive into the research topics of 'Neural Markovian predictive compression: An algorithm for online lossless data compression'. Together they form a unique fingerprint.

Cite this