On Information of Logical Expression and Knowledge Refinement
スポンサーリンク
概要
- 論文の詳細を見る
In machine learning, information theory has been recognized as a useful criterion, and several algorithms such as ID3 and Prism have been developed. In these methods, however, information theory is only used for generating inductively knowledge that is represented in decision tree or if then rule, but the issue on knowledge refinement is not considered. Moreover, they are not used for evaluating information of logical expression. This paper discusses a way of calculating quantitatively information of logical expression and its application for refining concept clusters discovered from a database. The calculation is based on the model representation of Multi-Layer Logic (MLL) with the hierarchical structure. Its key feature is the quantitative evaluation for selecting the best representation of the MLL formula by using cooperatively a criterion based on information theory (entropy) and domain knowledge.
- 一般社団法人情報処理学会の論文
- 1997-04-15
著者
-
ZHONG Ning
Department of Information Engineering, Maebashi Institute of Technology
-
Zhong N
Yamaguchi Univ.
-
Zhong Ning
Department Of Information Engineering
-
OHSUGA SETSUO
Department of Information and Computer Science, School of Science and Engineering, Waseda University
-
Ohsuga S
Waseda Univ.
-
Ohsuga Setsuo
Department Of Information And Computer Science School Of Science And Engineering Waseda University
-
Zhong Ning
Department Of Computer Science And Systems Engineering Faculty Of Engineering Yamaguchi University
関連論文
- Rough Sets in Knowledge Discovery and Data Mining (特集:ラフ集合の理論と応用)
- Soft Techniques for Rule Discovery in Data
- Organizing Multiple Data Sources for Developing Intelligent e-Business Portals
- On Information of Logical Expression and Knowledge Refinement
- A Multi-phase Process for Discovering, Managing, and Refining Strong Functional Relationships Hidden in Databases
- Effects of Chronic Intermittent Hypoxia on the Hemodynamics of Systemic Circulation in Rats