1998 Fiscal Year Final Research Report Summary
Learning Baysian Networks based on the MDL principle
| Project/Area Number |
09680367
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | Osaka University |
|---|
Principal Investigator |
SUZUKI Joe Osaka University Graduate school of Science Associate Professor, 大学院・理学研究科, 助教授 (50216397)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
SATAKE Ikuo Osaka University Graduate school of Science Research Associate, 大学院・理学研究科, 助手 (80243161)
KIKUCHI Kazunori Osaka University Graduate school of Science Research Associate, 大学院・理学研究科, 助手 (40252572)
TAKAHASHI Satoshi Osaka University Graduate school of Science Lecturer, 大学院・理学研究科, 講師 (70226835)
NAGATOMO Kiyokazu Osaka University Graduate school of Science Associate Professor, 大学院・理学研究科, 助教授 (90172543)
MURAKAMI Jun Osaka University Graduate school of Science Associate Professor, 大学院・理学研究科, 助教授 (90157751)
|
|---|
| Project Period (FY) |
1997 – 1998
|
| Keywords | MDL principle / Baysian Networks / machine learning / a prior knowledge / lranch and bound lechuzue |
|---|
| Research Abstract |
In this study, the computational issue in the problem of learning Bayesian belief networks (BBNs) based on the minimum description length (MDL) principle is addressed.Based on an asymptotic formula of description length, we apply the branch and bound technique to finding true network structures.The resulting algorithm searches considerably saves the computation yet successfully searches the network structure with the minimum value of the formula.Thus far, there has been no search algorithm that finds the optimal solution for examples of practical size and a set of network structures in the sense of the maximum posterior probability, and heuristic searches such as K2 and K3 trap in local optima due to the greedy nature even when the sample size is large.The proposed algorithm, since it minimizes the description length, eventually selects the true network structure as the sample size goes to infinity.
|
