An Evolving Associative Classifier for Incomplete Database

An associative classification method for incomplete database is proposed based on an evolutionary rule extraction method. The method can extract class association rules directly from the database including missing values and build an associative classifier. Instances including missing values are cla...

Full description

Saved in:
Bibliographic Details
Published inAdvances in Data Mining. Applications and Theoretical Aspects pp. 136 - 150
Main Author Shimada, Kaoru
Format Book Chapter
LanguageEnglish
Published Berlin, Heidelberg Springer Berlin Heidelberg
SeriesLecture Notes in Computer Science
Subjects
association rule
classification
evolutionary computation
incomplete data
Online AccessGet full text

Cover

More Information
Summary:An associative classification method for incomplete database is proposed based on an evolutionary rule extraction method. The method can extract class association rules directly from the database including missing values and build an associative classifier. Instances including missing values are classified by the classifier. In addition, an evolving associative classifier is proposed. The proposed method evolves the classifier using the labeled instances by itself as acquired information. The performance of the classification was evaluated using artificial incomplete data set. The results showed that the proposed evolving associative classifier has a potential to expand the target data for classification through its evolutionary process and gather useful information itself.
ISBN:3642314872
9783642314872
ISSN:0302-9743
1611-3349
DOI:10.1007/978-3-642-31488-9_12