A Novel Rough Fuzzy Clustering Algorithm with A New Similarity Measurement
Yang Li, Jiancong Fan, Jeng‐Shyang Pan, Gui-han Mao, Gengkun Wu · 網際網路技術學刊 · 2019
With the emergence of exponential growth of datasets in various fields, fuzzy theory-based approaches are widely used to improve or optimize the data clustering algorithms. These improved algorithms can achieve better results than the original counterparts in practical applications. However, the fuzzy clustering algorithms including the traditional improved algorithms normally ignore the clustering boundary uncertainty, inter-class compactness and complex data problems, thereby result in the unsatisfactory clustering results. To address this issue, in this paper, a novel rough fuzzy clustering algorithm based on a new similarity measure is proposed by utilizing the upper approximation and lower approximation of rough set. We also develop the method of transforming fuzzy clustering model into rough set model. Our experiment results show that the improved algorithm can get better clustering effect.