Abstract: For local optimum and initial sensitive problems with fuzzy C-means clustering, a new fuzzy clustering algorithm based on firefly algorithm is proposed. By incorporating the capacities of local and global search of firefly algorithm and FCM, taking the optimal clustering center of firefly algorithm as the initialized value of the FCM, and then clustering analysis is processed by FCM. The new algorithm overcomes FCM trapped local optimum and being sensitive to initial value effectively, and enhances the capacity of local search of firefly algorithm. The experimental results show that the new algorithm not only has better global search capacity and faster convergence speed but also has better clustering efficiency.

Key words: firefly algorithm, fuzzy clustering, fuzzy C-means clustering

摘要： 针对模糊C-均值聚类对初始值敏感、容易陷入局部最优的缺陷，提出了一种基于萤火虫算法的模糊聚类方法。该方法结合萤火虫算法良好的全局寻优能力和模糊C-均值算法的较强的局部搜索特性，用萤火虫算法优化搜索FCM的聚类中心，利用FCM进行聚类，有效地克服了模糊C-均值聚类的不足，同时增强了萤火虫算法的局部搜索能力。实验结果表明，该算法具有很好的全局寻优能力和较快的收敛速度，能有效地收敛于全局最优解，具有较好的聚类效果。

关键词: 萤火虫算法, 模糊聚类, 模糊C-均值聚类