Approximation of Bayes code for Markov sources

Jun ichi Takeuchi; Tsutomu Kawabata

Approximation of Bayes code for Markov sources

Research output: Contribution to conference › Paper › peer-review

Abstract

An approximation formula for the predictive Bayes code for the FSMX models, subspaces of Markov models, is presented. It is empirically shown that the code using the approximation formula along with the Jeffreys prior (Clark et al, 1994) gives shorter code length than the one using the Laplace estimator for the first order Markov models. It is also shown that there is difficulty in introducing the formula to CONTEXT or CTW methods.

Original language	English
Number of pages	1
Publication status	Published - Jan 1 1995
Externally published	Yes
Event	Proceedings of the 1995 IEEE International Symposium on Information Theory - Whistler, BC, Can Duration: Sept 17 1995 → Sept 22 1995

Other

Other	Proceedings of the 1995 IEEE International Symposium on Information Theory
City	Whistler, BC, Can
Period	9/17/95 → 9/22/95

All Science Journal Classification (ASJC) codes

Theoretical Computer Science
Information Systems
Modelling and Simulation
Applied Mathematics

Cite this

@conference{a2bfb4ce27bc4ba2b7e844f8d1a09b4d,

title = "Approximation of Bayes code for Markov sources",

abstract = "An approximation formula for the predictive Bayes code for the FSMX models, subspaces of Markov models, is presented. It is empirically shown that the code using the approximation formula along with the Jeffreys prior (Clark et al, 1994) gives shorter code length than the one using the Laplace estimator for the first order Markov models. It is also shown that there is difficulty in introducing the formula to CONTEXT or CTW methods.",

author = "Takeuchi, {Jun ichi} and Tsutomu Kawabata",

year = "1995",

month = jan,

day = "1",

language = "English",

note = "Proceedings of the 1995 IEEE International Symposium on Information Theory ; Conference date: 17-09-1995 Through 22-09-1995",

}

TY - CONF

T1 - Approximation of Bayes code for Markov sources

AU - Takeuchi, Jun ichi

AU - Kawabata, Tsutomu

PY - 1995/1/1

Y1 - 1995/1/1

N2 - An approximation formula for the predictive Bayes code for the FSMX models, subspaces of Markov models, is presented. It is empirically shown that the code using the approximation formula along with the Jeffreys prior (Clark et al, 1994) gives shorter code length than the one using the Laplace estimator for the first order Markov models. It is also shown that there is difficulty in introducing the formula to CONTEXT or CTW methods.

AB - An approximation formula for the predictive Bayes code for the FSMX models, subspaces of Markov models, is presented. It is empirically shown that the code using the approximation formula along with the Jeffreys prior (Clark et al, 1994) gives shorter code length than the one using the Laplace estimator for the first order Markov models. It is also shown that there is difficulty in introducing the formula to CONTEXT or CTW methods.

UR - http://www.scopus.com/inward/record.url?scp=0029192940&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0029192940&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0029192940

T2 - Proceedings of the 1995 IEEE International Symposium on Information Theory

Y2 - 17 September 1995 through 22 September 1995

ER -

Approximation of Bayes code for Markov sources

Abstract

Other

All Science Journal Classification (ASJC) codes

Other files and links

Fingerprint

Cite this