Accession number:20083311452476
Title:The characteristics of learning in limited data and the comparative assessment of learning methods
Authors:Chang, Fengming M. (1)
Author affiliation:(1) Department of Information Science and Applications, Asia University, Wufeng, Taichung 41354, Taiwan Corresponding author:Shang, F.M.
(paperss@gmail.com)
Source title:WSEAS Transactions on Information Science and Applications
Abbreviated source title:WSEAS Trans. Inf. Sci. Appl.
Volume:5 Issue:4
Issue date:April 2008 Publication year:2008 Pages:407-414
Language:English ISSN:17900832
Document type:Journal article (JA)
Publisher:World Scientific and Engineering Academy and Society, Ag.
Ioannou Theologou 17-23, Zographou, Athens, 15773, Greece Abstract:Many studies about learning in limited data were made in recent years. Without double, small data set learning is a
challenging problem. Information in data of small size is scarce and has some learning limit. While discussing the learning accuracy in limited data, different classification method causes different results for different data because each classification method has its
property. A method is the best solution for one data but is not the best for another. Therefore, this study analyzes the characteristics of small data set learning by the comparison of classification
methods. The Mega-fuzzification method for small data set learning is applied mainly. The comparison of different classification methods for small data set learning with several kinds of data is also
presented.
Number of references:21 Main heading:Education
Controlled terms:Classification (of information) - Learning systems
Uncontrolled terms:Best solution - Challenging problem - Classification method - Classification methods - Comparative
assessments - Learning accuracy - Learning methods - Limited data - Machine learning - Mega-fuzzification - Paucity of data - Small data set - Small data set learning - Small size
Classification code:461.4 Ergonomics and Human Factors
Engineering - 716.1 Information Theory and Signal Processing - 901.2 Education
Database:Compendex
Compilation and indexing terms, Copyright 2009 Elsevier Inc.
