Turkish Journal of Electrical Engineering and Computer Sciences
DOI
10.3906/elk-1008-726
Abstract
The goal of attribute reduction is to find a minimal subset (MS) R of the condition attribute set C of a dataset such that R has the same classification power as C. It was proved that the number of MSs for a dataset with n attributes may be as large as (_{n/2}^n) and the generation of all of them is an NP-hard problem. The main reason for this is the intractable space complexity of the conversion of the discernibility function (DF) of a dataset to the disjunctive normal form (DNF). Our analysis of many DF-to-DNF conversion processes showed that approximately (1-2/(_{n/2}^n) \times 100)% of the implicants generated in the DF-to-DNF process are redundant ones. We prevented their generation based on the Boolean inverse distribution law. Due to this property, the proposed method generates 0.5 \times (_{n/2}^n) times fewer implicants than other Boolean logic-based attribute reduction methods. Hence, it can process most of the datasets that cannot be processed by other attribute reduction methods.
Keywords
Information system, dataset, attribute reduction, feature selection, discernibility function, computational complexity, reduct
First Page
643
Last Page
656
Recommended Citation
HACIBEYOĞLU, MEHMET; BAŞÇİFTÇİ, FATİH; and KAHRAMANLI, ŞİRZAT
(2011)
"A logic method for efficient reduction of the space complexity of the attribute reduction problem,"
Turkish Journal of Electrical Engineering and Computer Sciences: Vol. 19:
No.
4, Article 10.
https://doi.org/10.3906/elk-1008-726
Available at:
https://journals.tubitak.gov.tr/elektrik/vol19/iss4/10
Included in
Computer Engineering Commons, Computer Sciences Commons, Electrical and Computer Engineering Commons