Hybrid subjective evaluation method using weighted subsethood - based (WSBA) rule generation algorithm

Fuzzy rules are important elements that should take into account in any fuzzy expert system.This paper proposes the framework of subjective performance evaluation using fuzzy technique for ranking the attributes of different types of datasets under a multi-criteria environment.The techniques such as...

Full description

Saved in:
Bibliographic Details
Main Authors: Othman, Mahmod, Khalid, Shaiful Annuar, Abdullah, Fader, Amir Hamzah, Shezrin Hawani, Ku-Mahamud, Ku Ruhana
Format: Article
Language:English
Published: ResearchersWorld 2013
Subjects:
Online Access:http://repo.uum.edu.my/9847/1/J.pdf
http://repo.uum.edu.my/9847/
http://www.researchersworld.com/vol4/index.html
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Fuzzy rules are important elements that should take into account in any fuzzy expert system.This paper proposes the framework of subjective performance evaluation using fuzzy technique for ranking the attributes of different types of datasets under a multi-criteria environment.The techniques such as fuzzy similarity function, fuzzy synthetic decision and satisfaction function have been adopted in these fuzzy evaluation methods.The framework is based on fuzzy multicriteria decision-making that consists of fuzzy rules. The use of fuzzy rules, which were extracted directly from input data through Weighted Subsethood-based (WSBA) Rule Generation Algorithm.WSBA rule generation use the subsethood values to generate the weights which finally produced the fuzzy general rules.The rules generated through the data provided knowledge in developed fuzzy rule The fuzzy rules embedded in the framework of subjective evaluation method showed advantages in generalizing the evaluation of the performance achievement, where the evaluation process can be conducted consistently in producing good evaluation results with the use of the membership set score.The results from the numerical examples are comparable to other fuzzy evaluation methods, even with the use of small rule size.