基于差分演化的K-均值聚类算法

传统的K-均值算法，因对初始聚类中心的选择敏感，存在容易陷入局部最优解的缺点,差分演化算法是一类基于种群的启发式全局搜索技术，对于实值参数的优化具有很强的鲁棒性。为了克服K－均值聚类算法的上述缺点，提出基于差分演化的K-均值聚类算法，该方法结合K－均值算法的高效性和差分演化算法的全局优化能力，较好地解决了聚类中心优化问题。通过实验结果表明，此算法能够有效改善聚类质量。

K-means Clustering Algorithm Based on Differential Evolution

The traditional K--means algorithm has the shortcoming that plunges into a local optimum prematurely because of sensitive selection of the initial cluster center; Differential evolution （DE） is a new heuristic global searching tech- nique based on population, which has been found to be very robust for real parameter＇s optimization. In order to overcome the shortcomings of K--means algorithm that mention above, proposed a K--means clustering algorithm based on DE. The algorithm proposed in this paper can well solve problem of optimizing cluster center by combining the high efficiency of K-- means algorithm with the ability of global optimization of DE. The experimental results show that algorithm proposed in this paper has improved the clustering quality effectively.