<abstract xmlns="http://www.w3.org/1999/xhtml"><p> Feature selection is a vital problem which needs to be effectively solved in knowledge discovery in databases and pattern recognition due to two basic reasons: minimizing costs and accurately classifying data. Feature selection using rough set theory is also called attribute reduction. It has attracted a lot of attention from researchers and numerous potential results have been gained. However, most of them are applied on static data and attribute reduction in dynamic databases is still in its early stages. This paper focuses on developing incremental methods and algorithms to derive reducts, employing a distance measure when decision systems vary in condition attribute set. We also conduct experiments on UCI data sets and the experimental results show that the proposed algorithms are better in terms of time consumption and reducts’ cardinality in comparison with non-incremental heuristic algorithm and the incremental approach using information entropy proposed by authors in [17].</p></abstract>

Feature selection is a vital problem which needs to be effectively solved in knowledge discovery in databases and pattern recognition due to two basic reasons: minimizing costs and accurately classifying data. Feature selection using rough set theory is also called attribute reduction. It has attracted a lot of attention from researchers and numerous potential results have been gained. However, most of them are applied on static data and attribute reduction in dynamic databases is still in its early stages. This paper focuses on developing incremental methods and algorithms to derive reducts, employing a distance measure when decision systems vary in condition attribute set. We also conduct experiments on UCI data sets and the experimental results show that the proposed algorithms are better in terms of time consumption and reducts’ cardinality in comparison with non-incremental heuristic algorithm and the incremental approach using information entropy proposed by authors in [17].

Metric Based Attribute Reduction Method in Dynamic Decision Tables

Institute for Computer and Control (SZTAKI), Hungarian Academy of Sciences, Hungary

Institute of Information Technology, VNU, Viet Nam

Institute of Information Technology, VAST, Viet Nam

Cybernetics and Information Technologies

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.

{"article-title":"Metric Based Attribute Reduction Method in Dynamic Decision Tables"}

Feature selection is a vital problem which needs to be effectively solved in knowledge discovery in databases and pattern recognition due to two basic reasons:...

Metric Based Attribute Reduction Method in Dynamic Decision Tables

Online veröffentlicht: 22. Juni 2016

Seitenbereich: 3 - 15

DOI: https://doi.org/10.1515/cait-2016-0016

Schlüsselwörter
Rough set, decision systems, attribute reduction, reduct, metric

© by Demetrovics Janos

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.

Metric Based Attribute Reduction Method in Dynamic Decision Tables

Online veröffentlicht: 22. Juni 2016

Seitenbereich: 3 - 15

DOI: https://doi.org/10.1515/cait-2016-0016

SchlüsselwörterRough set, decision systems, attribute reduction, reduct, metric

© by Demetrovics Janos

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.

Schlüsselwörter
Rough set, decision systems, attribute reduction, reduct, metric