Research On Fuzzy Clustering Based On Ordered Mechanism

Fuzzy C-Means clustering(FCM)is a representative algorithm in fuzzy clustering research.The algorithm is simple and efficient,but it is sensitive to noise data.Fuzzy C-Ordered-Means(FCOM)adds sorting operations,which reduces the influence of noise data on the centroid and enhances the robustness of the algorithm.However,on the one hand,the FCOM algorithm is difficult to process large-scale data,and on the other hand,the time complexity of the algorithm is high.Based on the FCOM algorithm,this paper studies the fuzzy clustering algorithm from two aspects of robustness and time complexity.The main research contents include:(1)In order to be able to process large-scale data,two fuzzy C-ordered mean clustering algorithms are proposed based on FCOM algorithm using Single-Pass and Online incremental framework: SPFCOM algorithm(Single-Pass Fuzzy C-OrderedMeans clustering)and OFCOM algorithm(Online Fuzzy C-Ordered-Means clustering).The SPFCOM and OFCOM algorithms inherit the ordered structure of the FCOM algorithm,which guarantees better cluster robustness and the ability to process large-scale data.In order to verify the correctness of the algorithm,this paper conducted experiments on several public data sets.The experimental results show that both SPFCOM and OFCOM algorithms are superior to the comparison algorithm in terms of clustering accuracy and robustness.(2)The FCOM algorithm uses the ordering method of ordered weighted average operators,which makes each sample have different influence on the cluster center point,which reduces the influence of noise data on clustering in the iterative process and improves the robustness of the algorithm.But the algorithm is more complex.In order to improve the clustering efficiency,a feature weighted Fuzzy C-Ordered-Means clustering(FWFCOM)algorithm is proposed based on the FCOM algorithm.The algorithm adds a method of range normalization,which simplifies the time-consuming sorting operation and improves the efficiency of the algorithm.This article uses multiple data sets for experiments.The results show that compared with the comparison algorithm,the FWFCOM algorithm effectively improves the efficiency of the algorithmwithout significantly affecting the accuracy and robustness of the cluster.