Simple Knowledge Reduction Approach Based on Direct Definitions and Discernibility Matrix

Reduction and core are two of the most important concepts in rough set theory. Two kinds of knowledge reduction methods are introduced based on the direct definitions and the discernibility matrix in this paper. A new knowledge reduction approach was proposed which combines the two aforementioned methods. The two steps of new method are: Firstly knowledge was classified and reduced with the indiscernibility relation, then core be calculated reusing distinction function. The experimental results show that the new method can simplify the calculation, and can reduce the computational complexity of the method.