Presentation 1999/7/16
On Bayes coding for word sequences
Takashi ISHIDA, Masayuki GOTOH, Shigeichi HIRASAWA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In universal source coding, the Ziv-Lempel (ZL) code and the Bayes code have been proposed. The improved types of the ZL code have been applied to data compression in practice. On the other hand, an efficient algorithm of the Bayes coding for FSMX model class has been proposed. The Bayes code has the good properties, such that asymptotic optimality and Bayes optimality. This implies that the Bayes code is superior to the ZL code for FSMX source from the viewpoint of compression rate. However, the fact such that the ZL code is often superior to the Bayes code for the practical data files have been reported. In this paper, we propose a new source model which emit word sequences and analyze the asymptotic performance of both codes for the proposed source models.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) suniversal coding / Bayes code / Ziv-Lempel code
Paper # IT99-27
Date of Issue

Conference Information
Committee IT
Conference Date 1999/7/16(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) On Bayes coding for word sequences
Sub Title (in English)
Keyword(1) suniversal coding
Keyword(2) Bayes code
Keyword(3) Ziv-Lempel code
1st Author's Name Takashi ISHIDA
1st Author's Affiliation Dep. of Industrial and Management Systems Engineering, Waseda University()
2nd Author's Name Masayuki GOTOH
2nd Author's Affiliation Dep. of Industrial and Management Systems Engineering, Waseda University
3rd Author's Name Shigeichi HIRASAWA
3rd Author's Affiliation Dep. of Industrial and Management Systems Engineering, Waseda University
Date 1999/7/16
Paper # IT99-27
Volume (vol) vol.99
Number (no) 187
Page pp.pp.-
#Pages 6
Date of Issue