基于超图正则化受限的概念分解算法.pdfVIP

第37 卷第3 期 电 子 与 信 息 学 报 Vol.37 No.3 2015 年3 月 Journal of Electronics Information Technology Mar. 2015 基于超图正则化受限的概念分解算法 李 雪* 赵春霞 舒振球 郭剑辉 (南京理工大学计算机科学与工程学院 南京 210094) 摘 要：针对概念分解(Concept Factorization, CF)算法没有同时考虑样本中存在的类别信息及数据间多元几何结 构信息的问题，该文提出一种基于超图正则化受限的概念分解(Hyper-graph regularized Constrained Concept Factorization, HCCF)算法。HCCF 算法通过构建一个无向加权的拉普拉斯超图正则项，提取数据间的多元几何结 构信息，克服了传统图模型只能表达数据间成对关系的缺陷；同时采用硬约束的方式使样本的类别信息在低维空间 中保持一致，充分利用了标记样本的类别信息。该文采用乘性迭代的方法求解HCCF 算法的目标函数并证明了其 收敛性。在TDT2 库、Reuters 库和PIE 库上的实验结果表明，HCCF 算法提高了聚类的准确率和归一化互信息， 验证了算法的有效性。 关键词：信息处理；概念分解；聚类；硬约束；超图；流形学习 中图分类号：TP391 文献标识码：A 文章编号：1009-5896(2015)03-0509-07 DOI: 10.11999/JEIT140799 Hyper-graph Regularized Constrained Concept Factorization Algorithm Li Xue Zhao Chun-xia Shu Zhen-qiu Guo Jian-hui (College of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing 210094, China) Abstract: The Concept Factorization (CF) algorithm can not take into account the label information and the multi-relationship of samples simultaneously. In this paper, a novel algorithm called Hyper-graph regularized Constrained Concept Factorization (HCCF) is proposed, which extracts the multi-geometry information of samples by constructing an undirected weighted hyper-graph Laplacian regularize term, hence overcomes the deficiency that traditional graph model expresses pair-wise relationship only. Meanwhile, HCCF takes full advantage of the label information of labeled samples as hard constraints, and it preserves label consistent in low-dimensional space. The objective function of HCCF is solved by the iterative multiplicative updating a