Presentation 2006-07-27
A Study of Bayes Coding for i.i.d. Sources with Consideration of the Generating Patterns of the Symbols in the Source Alphabet
Ryunosuke NANMO, Daiki KOIZUMI, Toshiyasu MATSUSHIMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Bayes code is one of the universal coding methods under the condition that the class of the probability distribution of the source is known but its parameters are unknown. Bayes code can minimize the redundancy in terms of Bayes decision theory. In conventional researches of universal coding, they consider only the source that generates all the symbols in the alphabet. The files in the real world, however, are not always regarded as this specific source. Therefore, it is crucial to deal with universal coding method under the condition that the generating patterns of the symbols are unknown. There are some researches from this point of view, however, they have not defined the source model clearly. By considering this point, this paper defines the source model when the generating patterns of the symbols are unknown and also defines Bayes code under this generalized source. Furthermore, the asymptotic code length is evaluated and the effective algorithm with some restrictions on the prior distribution of patterns of symbols is also proposed.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Bayes code / universal coding / redundancy / prior distribution
Paper # IT2006-29
Date of Issue

Conference Information
Committee IT
Conference Date 2006/7/20(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Information Theory (IT)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Study of Bayes Coding for i.i.d. Sources with Consideration of the Generating Patterns of the Symbols in the Source Alphabet
Sub Title (in English)
Keyword(1) Bayes code
Keyword(2) universal coding
Keyword(3) redundancy
Keyword(4) prior distribution
1st Author's Name Ryunosuke NANMO
1st Author's Affiliation Department of Industrial Management Systems Engineering, School of Science and Engineering, Waseda University()
2nd Author's Name Daiki KOIZUMI
2nd Author's Affiliation Department of Industrial Management Systems Engineering, School of Science and Engineering, Waseda University
3rd Author's Name Toshiyasu MATSUSHIMA
3rd Author's Affiliation Department of Industrial Management Systems Engineering, School of Science and Engineering, Waseda University
Date 2006-07-27
Paper # IT2006-29
Volume (vol) vol.106
Number (no) 184
Page pp.pp.-
#Pages 6
Date of Issue