Current Proceedings on Technology
Yazarlar: Rabiei Mamat, Tutut Herawan, Mustafa Mat Deris
Konular:-
Anahtar Kelimeler:Data Clustering,Decision Attribute Selection,Soft Set Theory,
Özet: Many rough set based techniques for partition attribute selection has been proposed. However, these techniques still requires high computational costs when involving large datasets. In this paper, a technique for selecting partition attribute called SAR (Super Attribute Representative) which is based on the soft set theory is proposed. Using this technique, comparisons at the attribute level is firstly computed before representative for each attributes compared to determine the partition attribute. Results show that on this technique, execution time is slightly reduces as compare to the rough set-based techniques.