Title :
Finding Value Reducts Using Rough Multisets
Author :
Chan, Chien-Chung ; Chen, Chen ; Garikapati, Varundev
Author_Institution :
Dept. of Comput. Sci., Univ. of Akron, Akron, OH, USA
Abstract :
This paper introduces a new value reduction algorithm for generating rules from consistent and inconsistent examples represented by Multiset Decision Tables (MDT). The algorithm is implemented as stored procedure in MS SQL, hence, it is capable of dealing with very large data sets. The running time of value reduction algorithm is higher than the decision tree program, in general. However, by using a minimum support threshold, the run-time can be improved significantly as indicated in our experiments using the IDS data set of 4 million records.
Keywords :
SQL; data reduction; decision tables; decision trees; rough set theory; IDS data; MS SQL; decision tree program; multiset decision tables; rough multisets; value reduction algorithm; Approximation methods; Decision trees; Finite element methods; Information systems; Partitioning algorithms; Rough sets;
Conference_Titel :
Granular Computing (GrC), 2010 IEEE International Conference on
Conference_Location :
San Jose, CA
Print_ISBN :
978-1-4244-7964-1
DOI :
10.1109/GrC.2010.150
