An alternative approach to normal parameter reduction algorithms for decision making using a soft set theory / Sani Danjuma

摘要

The soft set theory is a mathematical tool that deals with uncertainty, imprecise and vagueness in decision systems. It has been widely used to identify irrelevant parameters and make reduction of parameters for decision making, in order to bring out the optimal choices of the decision systems. Many normal parameter reduction algorithms exist to handle parameter reduction and maintain consistency of decision choices. However, they require much time to repeatedly run the algorithms to reduce unnecessary parameters using either parameter important degree or oriented parameter sum. This study will firstly review the different parameter reduction and decision making techniques for soft set and hybrid soft sets under unpleasant set of hypothesis environment as well as performance analysis of their derived algorithms. Consequently, the summary of the current literature in those areas of research were given, pointed out the limitations of previous works and areas that require further research works. Secondly, an alternative algorithm for parameter reduction and decision making based on soft set theory was proposed. The proposed algorithm showed that it can reduce the computational complexity and run time compared to baseline algorithms. Finally, to evaluate the proposed algorithm, thorough experimentation on both real life and synthetic binary-valued data set were performed. The experimental result shows that the proposed algorithm was feasible and has relatively reduced the computational complexity and running time with an average of 56 percent compared with the existing algorithms. In addition, the algorithm was relatively easy to understand compare to the state of the art of normal parameter reduction algorithm. The proposed algorithm was able to avoid the use of parameter important degree, decision partition and finding the multiple of the universe within the sets. This study contributes significantly in reducing the computational complexity and running time as compared with Normal Parameter Reduction algorithm (NPR) and New Efficient Normal Parameter Reduction algorithm (NENPR).