Research On Hesitant Fuzzy Clustering Algorithm And Complexity Optimization

With the continuous development of the information era,data is not just deterministic data and uncertain data such as tuples and attribute values.The proportion of "indecision" data that accords with human character is increasing.Such data is inherently uncertain,so it cannot be simply described with a certain value or the probability that a certain value exists.The hesitant fuzzy set is suitable for solving this kind of hesitant fuzzy problem.Today,the hesitant fuzzy set correlation clustering algorithm has high research value.In this paper,the following research work is done to solve some problems of the current hesitant fuzzy clustering algorithms:1.Unified the concept definition of hesitant fuzzy clustering.In the existing related researches,some scholars only provided language descriptions for the relevant definitions of hesitant fuzzy sets,but did not give formal definitions and mathematical expressions;some went further and gave definitions of their mathematical expressions,but this definition is not convenient for solving the clustering problem.In view of this,this article will uniformly describe the mathematical symbols and mathematical formulas in order to effectively solve the clustering problem.2.This paper proposes a fuzzy hierarchical clustering algorithm with constant hesitancy of agglomeration centers.Aiming at the problems of high time complexity and high space complexity in the original hesitant fuzzy hierarchical clustering algorithm,and the algorithm may not be able to obtain ideal results when solving real-life problems.This algorithm uses a new weight formula and an updated formula for the cluster center.In conjunction with the formal definition of the hesitant fuzzy clustering concept,the algorithm guarantees linear changes in time complexity and space complexity during the iterative process.At the same time,the quality of clustering is not lower than the original algorithm,which can effectively solve the problem of hesitant fuzzy data clustering in real life.3.This paper proposes a fuzzy hierarchical clustering algorithm with constant hesitancy of agglomeration centers.Aiming at the problems of high time complexity and high space complexity in the original hesitant fuzzy hierarchical clustering algorithm,and the algorithm may not be able to obtain ideal results when solvingreal-life problems.This algorithm uses a new weight formula and an updated formula for the cluster center.In conjunction with the formal definition of the hesitant fuzzy clustering concept,the algorithm guarantees linear changes in time complexity and space complexity during the iterative process.At the same time,the quality of clustering is not lower than the original algorithm,which can effectively solve the problem of hesitant fuzzy data clustering in real life.4.A fuzzy hierarchical clustering algorithm with adaptive hesitation is proposed.The algorithm is extended and optimized on the basis of the fuzzy hierarchical clustering algorithm with constant hesitancy of the aggregation center.On the basis of reducing the time and space complexity of the original algorithm,the calculation formula of the cluster center was redefined,so that the hesitation degree of the cluster center can be automatically determined,and the data of the original hesitant fuzzy object set is retained to the greatest extent.Information has better robustness when dealing with hesitant fuzzy sets with larger hesitation.