A New Approach for Attribute Reduction from Decision Table based on Intuitionistic Fuzzy Topology

Abstract Most of the current attribute reduction methods use the measure to define the reduct, such as positive region of rough set theory (RST), granular information entropy, and granular distance measures. However, the reducts defined based on the measures are still limited in sizes because of the information conservation property of the reduct compared to the original dataset. Dealing with this matter, this paper proposes a new approach using the intuitionistic fuzzy topology (IFT) to develop an attribute reduction method to find the reduct with the optimal size through topology. Specifically, the IFT base and IFT sub-base structures are proposed based on the pre-order relation. Secondly, we propose a new measure to evaluate the significance of the attribute based on the IFT sub-base. Lastly, we propose a new definition of the reduct based on the IFT base and design two new algorithms based on Heuristic approaches. The theoretical and experimental results show that the proposed method is efficient in size and accuracy of the reduct