2006 Fiscal Year Final Research Report Summary
Computer-oriented analysis of incomplete data
| Project/Area Number |
16300087
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Statistical science
|
|---|
| Research Institution | RIKKYO UNIVERSITY |
|---|
Principal Investigator |
YAMAGUCHI Kazunori Rikkyo University, College of Business, Professor, 経営学部, 教授 (60230348)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
TAKEUCHI Akinobu Jissen women's, Department of Humanities and Social Sciences, Lecturer, 人間社会学部, 講師 (60339596)
WATANABE Michiko Toyo University, Faculty of Economic, Professor, 経済学部, 教授 (50150397)
NAKAGAWA Sigekazu Kurashiki University of Science and the Arts, Associate Professor, 産業科学技術学部, 助教授 (90248203)
KURODA Masahiro Okayama University of Science, Faculty of Informatics, Lecturer, 総合情報学部, 講師 (90279042)
SAKAORI Fumitake Rikkyo University, College of Social Relations, Research, Associate, 社会学部, 助手 (90386475)
|
|---|
| Project Period (FY) |
2004 – 2006
|
| Keywords | EM algorithm / Propensit score / Bayes analysis / Latent class model |
|---|
| Research Abstract |
For the analysis of incomplete data, in particular, data with missing values, this project has had the following results;
(1) Latent class models can be used for the covariate adjustment under missing at random conditions
See Sakaori and Yamaguchi (2006).
(2) For the analysis of multivariate count data with missing values, latent class modes with several latent variables have been proposed.
See Yamaguchi, Sakurai, and Watanabe (2006).
(3) An accelerating method for the EM algorithm has been proposed,
See Kuroda and Sakakibara (2006).
|
