Research And Applications Of Fuzzy Clustering Based On Flower Pollination Algorithm

Data mining is a very important area that is booming in the current society.Clustering analysis is a hot issue in data mining.As a branch of statistics,Clustering analysis relies on statistical methods,machine learning and neural network.The fuzzy clustering algorithm applies the basic theory of fuzzy set to the clustering analysis,which provide some fuzzy processing ability for the displayed data.The Fuzzy cmeans algorithm is a very important research method in the fuzzy clustering.It is obvious that the FCM algorithm also lead to local minimum and depends on initial value.We have deeply studied the FCM algorithm and flower pollination(FPA)algorithm,so we proposed a new fuzzy clustering based on flower pollination algorithm(FPAFCM).The main research as follows:(1)Explained the basic principles of flower pollination algorithm and fuzzy clustering algorithm.At the same time,we also did the research at home and abroad,and analyzed the advantages and disadvantages of flower pollination algorithm and fuzzy clustering algorithm this paper.(2)We improve the standard flower pollination algorithm,and put forward the collaborative search strategy into the flower pollination algorithm.Through the exchange and cooperation among the populations,we update the historical global optimal value.Through several test,we choose eight benchmark test functions.The COFPA algorithm,the FPA algorithm and the improved algorithm are repeatedly tested on eight benchmark test functions respectively.Experiments show that the COFPA algorithm has obvious improvement than the FPA algorithm and the improved flower pollination algorithm.(3)By combining the improved flower pollination algorithm with the fuzzy clustering algorithm,we proposed a new fuzzy clustering algorithm based on flower pollination(FPAFCM)algorithm.The FPAFCM algorithm was applied to three UCI standard data sets.We compared the FPAFCM algorithm and the reference.Experiments show that the accuracy of FPAFCM algorithm is better than FCM algorithm and the improved algorithm.(4)Summarize the research results and propose further research questions.