【摘要】 文中提出一种半监督核信任力传播聚类算法(SSKAPC).SSKAPC在对样本聚类的过程中,引入先验知识提高聚类性能;同时该算法将样本映射到高维空间进行聚类.人工数据和真实世界数据的实验表明,SSKAPC算法能大幅度提高聚类的准确性.更多还原

【Abstract】 In this paper,a semi-supervised kernel affinity propagation method named SSKAPC is proposed.In this method,affinity propagation clustering method is extended to semi-supervised setting,in which background knowledge is provided in terms of pairwise constraints.Kernel trick is also used to process non-linear problem.The experimental results and comparisons on simulated and real-world datasets illustrate the effectiveness and the advantages of the proposed SSKAPC method.更多还原