Improvement And Application Of Particle Swarm Optimization Algorithm

With the continuous development of information technology and data storage technology,the data mining technology becomes one of the most active research directions in the field of information decision.As a powerful analytical tool for data mining,clustering has obtained widespread concern.Traditional clustering methods were difficult to deal with large-scale,complex data.In this case,some modern intelligent clustering algorithms have emerged,among which Particle Swarm Optimization algorithm is excellent in dealing with the problems of large-scale data.Particle swarm optimization clustering is a combination of the data clustering analysis and the biological mechanism of the particle swarm.Has the ability of self-organizing and self-learning have access to strong adaptability and robustness,which make it can get the global optimal solution.First of all,the paper introduces the particle swarm clustering algorithm research background and significance,as well as domestic and foreign research status;and then gives the particle swarm algorithm,clustering algorithm and particle swarm clustering algorithm based on the basic theory and methods.Then focuses on the improved algorithm of particle swarm clustering algorithm.The algorithm introduces cluster validity index to effective measure the clustering results,and through the objective function of the PSO clustering algorithm,the cluster validity index and the particle swarm's crossover and mutation to improve the efficiency and accuracy of clustering algorithm.At last,using the improved PSO clustering algorithm to do three small application:first of all cluster analysis in the simulated data sets;second doing a certain application in medical image segmentation,finally doing some application in the intelligent traffic on the shortest path selection.The results of these applications show that the improved PSO clustering algorithm has better clustering effect,higher accuracy,stronger robustness and higher efficiency.The proposed algorithm has a strong applicability.