Study On Clustering Problem Based On Improved Cuckoo Algorithm

With the advent of the big data era and the development of cloud computing technology,data has exploded.Faced with such a large number of complicated data,how to mine valuable information from these massive data is a problem that we must solve.Clustering is one of the most important research directions in the data mining system.It has a wide range of applications in many fields.Selecting reasonable problem models and methods can effectively obtain the required effective information from massive data and make use of it.In this paper,the capacity constrained clustering model(CCP)is studied,which is a typical classical clustering location problem.Aiming at the shortcoming that the original clustering method depends on the initial solution and is easy to fall into local optimization,this paper proposes an improved cuckoo search algorithm to solve the problem.First of all,the Cuckoo search algorithm(CS)has strong global search capability and interaction capability.It can escape from the local optimal trap through Levy's flight update mechanism,thus improving the diversity of the population and having certain advantages.Secondly,on the basis of this algorithm,in order to improve the search efficiency of the CS algorithm,K-Means is used to determine the initial center and quickly generate the initial solution.At the same time,the cycle transfer(CT)is used to further optimize the solution.Finding feasible negative rings in data sets,expanding the spatial search range of solutions and improving the quality of solutions can obtain the approximate optimal solution of the problem more quickly.In this paper,the data set in OR-Library is used as the example data source for experimental comparison,and the calculation program is written by MATLAB language.The results show that on the standard data set used in the experiment,firstly,compared with the original K-Means algorithm and Cuckoo algorithm,the performan ce of the improved Cuckoo algorithm is obviously improved.Secondly,compared with the optimal value of the standard data set,the improved cuckoo algorithm is closer to the optimal value of the data set than the other three heuristic algorithms and has advantages.Finally,the actual data of logistics distribution center location are used to verify,and the experimental results of the improved algorithm proposed in this paper are compared and analyzed with the algorithms in the literature.The experimental results show that the improved algorithm in this paper has a good effect on solving clustering problems,which shows that the method proposed in this paper has certain feasibility and superiority.