| Presentation | 2004/3/9 A Study on Context Tree Weighting Method for Non-Stationary Source Tatsuya HISATOMI, Masazumi KURIHARA, Kazuhiko YAMAGUCHI, Kingo KOBAYASHI, |
|---|---|
| PDF Download Page | PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | It is well known that Bayes code attains the minimum redundancy when the information source is assumed to belong in a class with some kind of regular conditions, but the paramenters characterized the actual source are unknown. The context tree weighting method [1] is an algorithm for Bayes code for FMSX sources of any depth. If we want to apply the the context tree weighting method for a class of nonstationary sources, it is necessary to throw out some amounts of past data from the context tree in order to eliminate the undesirable effect degrading the performance of parameter estimation. In this paper, we propose a method adapting for the change of states of nonstationary source. The method continuously estimates the nonstationary source parameters at each node of the context tree weighting method, and can adapt to FSMX sources of any depth compared with the conventional FWCTW method [2]. Finally we show experimental results that indicate the convergence of the codeword length of our method to the ideal one of CTW method. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Universal source coding / Non-stationary source / Bayes coding / Context Tree Weighting Method |
| Paper # | IT2003-73,ISEC2003-113,WBS2003-191 |
| Date of Issue |
| Conference Information | |
| Committee | ISEC |
|---|---|
| Conference Date | 2004/3/9(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 Security (ISEC) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | A Study on Context Tree Weighting Method for Non-Stationary Source |
| Sub Title (in English) | |
| Keyword(1) | Universal source coding |
| Keyword(2) | Non-stationary source |
| Keyword(3) | Bayes coding |
| Keyword(4) | Context Tree Weighting Method |
| 1st Author's Name | Tatsuya HISATOMI |
| 1st Author's Affiliation | Dept. of Information Engineering, The Graduate School of Electro-Communications, The University of Electro-Communications() |
| 2nd Author's Name | Masazumi KURIHARA |
| 2nd Author's Affiliation | Dept. of Information and Communication Engineering, The Graduate School of Electro-Communications, The University of Electro-Communications |
| 3rd Author's Name | Kazuhiko YAMAGUCHI |
| 3rd Author's Affiliation | Dept. of Information and Communication Engineering, The Graduate School of Electro-Communications, The University of Electro-Communications |
| 4th Author's Name | Kingo KOBAYASHI |
| 4th Author's Affiliation | Dept. of Information and Communication Engineering, The Graduate School of Electro-Communications, The University of Electro-Communications |
| Date | 2004/3/9 |
| Paper # | IT2003-73,ISEC2003-113,WBS2003-191 |
| Volume (vol) | vol.103 |
| Number (no) | 713 |
| Page | pp.pp.- |
| #Pages | 5 |
| Date of Issue | |