Efficient hybrid reduction for binary based information system in soft set theory

• Main Author: Mohd Rose, Ahmad Nazari
• Format: Thesis
• Language: English; English; English
• Published: 2016
• Subjects: QA Mathematics; QA150-272.5 Algebra
• Online Access: http://eprints.uthm.edu.my/9997/2/24p%20AHMAD%20NAZARI%20MOHD%20ROSE.pdf; http://eprints.uthm.edu.my/9997/1/AHMAD%20NAZARI%20MOHD%20ROSE%20COPYRIGHT%20DECLARATION.pdf; http://eprints.uthm.edu.my/9997/3/AHMAD%20NAZARI%20MOHD%20ROSE%20WATERMARK.pdf; http://eprints.uthm.edu.my/9997/
• Institution: Universiti Tun Hussein Onn Malaysia

In soft set literatures, issues regarding reduction techniques with regards to dataset in soft set have been discussed and analyzed. The existing reduction techniques discussed were the techniques based on rough set guidelines and parameter reduction. All of the proposed techniques have successfully reduced the datasets but the factors of consistency and accuracy are still outstanding. Based on the research done oi:i data transformation in soft set theory, the three newly introduced reduction methods will be integrated into a technique known as Hybrid Reduction in Soft Set (HRSS). HRSS consists of two(2) types of parameter reduction and a newly proposed object reduction. The proposed technique has been implemented and the results were compared to the existing techniques, and HRSS was found to be I 00% consistent, accurate and able to reduce the data substantially. With SRR (Soft Set Rough Reduction) and Parameter Reduction (PR) being ineffective with respect to consistency and accuracy, further analysis on the data size achieved by HRSS and Normal Parameter Reduction (NPR) were then considered. HRSS has also demonstrated efficiency when searching for decisional values. Lastly, HRSS has also been found to be the least complexed in terms of the algorithm used. With the results obtained, it is safe to conclude that, decision-making that are based on selected datasets that have undergone the HRSS processing is competent