Consistency of Bayesian Network Structure Learning
| Project/Area Number |
21500138
|
|---|
| 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 大阪大学, 大学院・理学研究科, 准教授 (50216397)
|
|---|
| Co-Investigator(Renkei-kenkyūsha) |
FUJIWARA Akio 大阪大学, 大学院・理学研究科, 教授 (30251359)
|
|---|
| Project Period (FY) |
2009 – 2012
|
| Project Status |
Completed (Fiscal Year 2012)
|
| Budget Amount *help |
¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000)
Fiscal Year 2011: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2010: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2009: ¥2,080,000 (Direct Cost: ¥1,600,000、Indirect Cost: ¥480,000)
|
|---|
| Keywords | ベイズネット / 情報量基準 / 一致性 / ガウシアン / 統計科学 / ベイジアンネットワーク / モデル選択 / 因果同定 / 離散変数と連続変数 / 線形回帰分析 / Hannan-Quinn |
|---|
| Research Abstract |
This research address the problem of finding the true structure of a Baysian network (BN) from a finite sample of examples. In particular, we apply information criteria which is expressed by the sum of likelihood and complexity of the BN structure. If the latter is too stressed, we cannot obtain the true structure even if the number of examples grows. We found the condition that the true one is obtained with probability one, and proved it.
|
Report
(4 results)
Research Products
(57 results)
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
[Presentation]2011
Author(s)
T. Inazumi, T. Washio, S. Shimizu, J.Suzuki, A. Yamamoto and Y. Kawahara
Organizer
Proc.UAI2011:The 27th Conf. on Uncertainty inArtificial Intelligence
Year and Date
2011-07-21
Related Report
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
