Project/Area Number |
10143104
|
Research Category |
Grant-in-Aid for Scientific Research on Priority Areas (A)
|
Allocation Type | Single-year Grants |
Research Institution | Tokyo Institute of Technology |
Principal Investigator |
SATO Taisuke Graduate School of Information Science and Engineering, Tokyo Institute of Technology, Prof., 大学院・情報理工学研究科, 教授 (90272690)
|
Co-Investigator(Kenkyū-buntansha) |
HARAGUCHI Makoto Hokkaido Univ., Graduate School of Engineering, Prof., 大学院・工学研究科, 教授 (40128450)
IMAI Mutsumi Keio Univ., Faculty Environmental Information, Assistant Prof., 環境情報学部, 助教授 (60255601)
ARIMURA Hiroki Kyushu Univ., Graduate School of Information Science and Electrical Engineering, Assistant Prof., 大学院・システム情報科学研究科, 助教授 (20222763)
SATO Masako Osaka Prefecture Univ., College of Integrated Arts and Sciences, Prof., 総合科学部, 教授 (50081419)
SHINOHARA Takeshi Kyushu Institute of Technology, Faculty of Computer Science and Systems Engineering, Prof., 情報工学部, 教授 (60154225)
古川 康一 慶應義塾大学, 大学院・政策・メディア研究科, 教授 (10245615)
|
Project Period (FY) |
1998 – 2000
|
Project Status |
Completed (Fiscal Year 2001)
|
Budget Amount *help |
¥85,700,000 (Direct Cost: ¥85,700,000)
Fiscal Year 2000: ¥28,300,000 (Direct Cost: ¥28,300,000)
Fiscal Year 1999: ¥21,600,000 (Direct Cost: ¥21,600,000)
Fiscal Year 1998: ¥35,800,000 (Direct Cost: ¥35,800,000)
|
Keywords | knowledge discovery / inductive inference / abduction / ILP / 記号的統計モデリング / 脳機能画像 / 拡張アブダクション / 高速パターンマッチング / 事例ベース / 節管理 / 抽象値 / 統計的記号モデリング / データマイニングアルゴリズム / H-map / 決定木 / 反駁推論 / 発見 |
Research Abstract |
From the fiscal year 1999 to 2001, we conducted research in knowledge discovery and developed inference various methods that can cope with uncertainty and complexities in the real data as follows. T. Sato developed a symbolic-statistical modeling language PRISM. Arimura et al. developed two types of fast text pattern matching algorithm. Tsukimoto proposed logical regression analysis. K.Satoh continued the analysis of minimal case base required for representing concepts. Sakama proposed non-monotonic inverse resolution. Imai et al. developed an algorithm for computing maximally specific hypotheses for ILP. Haraguchi proposed data abstraction for decision trees. Yamamoto reconstructed the theoretical base of ILP based on the logic of Suggestion. M. Sato et al. studied the theory of refutable/inductive inference. Shinohara proposed reducing data dimension by his H-map. Ohsawa showed that his KeyGraph approach is effective by examples.
|