A Classification Approach Using Improved Neighborhood Rough Set Feature Selection and Fuzzy K-Nearest Neighbor Method
Chengfeng Zheng, Zhizhong Yan, Mohd Shareduwan Mohd Kasihmuddin, Chunqiu Wei, Mohd. Asyraf Mansor · 2023
This paper proposes a classification methodology that combines neighborhood rough set feature selection with the fuzzy k-nearest neighbor algorithm. The method enhances the accuracy by integrating the fuzzy k-nearest neighbor classifier into the well-established principles of the improved neighborhood rough set feature selection. Through rigorous experimentation, we applied the proposed classification approach to numerically simulate diverse datasets across various application scenarios. The obtained results consistently showcase the high numerical performance achieved by the method. Furthermore, a comprehensive comparative analysis with baseline methods further validates the effectiveness of our approach.