基于遗传算法的直觉模糊C均值聚类算法

Intuitionistic Fuzzy C-means Clustering Algorithms Based on Genetic Algorithms

针对一般直觉模糊C均值聚类算法在寻优过程中易陷入局部最优解的问题,利用遗传算法具备全局寻优的优点,提出了一种基于遗传算法的直觉模糊C均值聚类算法。在该算法中聚类中心为直觉模糊数,这使得遗传过程中个体信息变得复杂,进而增大了约束问题的处理难度。本文对产生的个体采用适时分段的归一化方法,很好地解决了该问题。仿真结果表明该算法所得聚类结果不仅准确而且更为细致。

In this paper, an intuitionistic fuzzy C-means clustering algorithms （IFCM） based on genetic algorithms is proposed. Compared with other methods based on various similarity matrixs, more objective results can be obtained by utilizing the optimal method to do clustering analysis. Firstly, the IFCM clustering method currently in use is discussed. By using this method, a local optimal value may be obtained as fuzzy C-means. The method proposed in this paper can overcome this drawback by combining that method with the GA method. In this process, the main problem is that the real number is changed into the intuitionistic fuzzy number. As a result, the clustering center is also changed from a real number to an intuitionistic fuzzy number. There may be some difficulty in handling with each clustering center, because the sum of membership, non-membership and uncertainty must be considered for each intuitionistic fuzzy number. In the GA program, after the cross operation, normalization is done for each clustering center to get the ideal results. At the end of this paper, an example is given and three methods are compared to illustrate the effectiveness of this method.