基于C-均值和免疫遗传算法的聚类分析

Cluster Analysis Based on C-Means and Immune Genetic Algorithm

聚类问题在一定条件下可以归结为一个带约束的优化问题。遗传算法作为一种鲁棒性很强的优化算法，具有很强的全局寻优能力。提出了一种基于C-均值和带免疫机制的混合遗传算法。理论分析和仿真实验表明，该算法既具有很强的全局寻优能力，也具有较强的局部寻优能力。

Cluster analysis is a kind of unsupervised learning method, which can extract the hidden rules from the feature data set of the objects. Clustering can be regarded as a constrained optimization problem under certain conditions. As a robust optimizing method, genetic algorithm has shown great global searching capability, which is independent of the problem domain. This paper proposes an improved hybrid genetic algorithm based on C-means and immune principle. Theoretical analysis and experiments show that this method outperforms the existing genetic clustering algorithms in both global and local convergence speed.