Cited 0 time in webofscience Cited 0 time in scopus

Compression By and For Deep Boltzmann Machines

Title
Compression By and For Deep Boltzmann Machines
Authors
Li, QingChen, YangKim, Yongjune
DGIST Authors
Kim, Yongjune
Issue Date
2020-12
Citation
IEEE Transactions on Communications, 68(12), 7498-7510
Type
Article
Article Type
Article
Author Keywords
Image codingRate-distortionNoise reductionMachine learningTrainingDistortionSource codingRate distortionlossy source codingdeep Boltzmann machinesdeep auto-encoderBlahut-Arimoto algorithmdenoising
Keywords
CAPACITYCODESREPRESENTATIONSINFORMATION
ISSN
0090-6778
Abstract
We answer two questions in this work: what Deep Boltzmann Machines (DBMs) can do for compression and vise versa. We show that (1) DBMs can be applied to learn the rate distortion approaching posterior as in the Blahut-Arimoto (BA) algorithm, and to construct a lossy source compression scheme based on the Deep AutoEncoder; (2) compression can improve DBMs' training performances via compression-based denoising algorithms. The implementation of the BA algorithm in the form of DBMs is the foundation of the two applications. © 1972-2012 IEEE.
URI
http://hdl.handle.net/20.500.11750/12484
DOI
10.1109/tcomm.2020.3020796
Publisher
Institute of Electrical and Electronics Engineers Inc.
Related Researcher
  • Author Kim, Yongjune Information, Computing, and Intelligence Laboratory
  • Research Interests machine learning, artificial intelligence, information theory, coding theory, storage, computing
Files:
There are no files associated with this item.
Collection:
Department of Information and Communication EngineeringInformation, Computing, and Intelligence Laboratory1. Journal Articles


qrcode mendeley

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE