Improved method of discernibility matrix and attributes reduction
Huang Chun-nan
Abstract
Huang Chun-nan
Abstract
With a view to eliminating too much calculation and greater complexity due to current method of attribute reduction,this paper proposes a new definition of discernibility matrix,based on discernibility matrix and the method of computing core,and according to the special quality of binary decision table.The improved method of forming discernibility matrix involves changing a complex matrix into two simple ones and gives fewer times of elements comparison than only one matrix.The analysis of the given data shows that the method improves the definition of discernibility matrix with accompanying simplification of the algorithm and enhancement of calculation efficiency.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
With a view to eliminating too much calculation and greater complexity due to current method of attribute reduction,this paper proposes a new definition of discernibility matrix,based on discernibility matrix and the method of computing core,and according to the special quality of binary decision table.The improved method of forming discernibility matrix involves changing a complex matrix into two simple ones and gives fewer times of elements comparison than only one matrix.The analysis of the given data shows that the method improves the definition of discernibility matrix with accompanying simplification of the algorithm and enhancement of calculation efficiency.
Key concepts: Matrix (chemical analysis), Reduction (mathematics), Decision table, Simple (philosophy), Mathematics, Rough set, Matrix method, Algorithm