《计算机应用》唯一官方网站 ›› 2022, Vol. 42 ›› Issue (1): 115-122.DOI: 10.11772/j.issn.1001-9081.2021071181

• 数据科学与技术 • 上一篇    

基于Hessian正则化和非负约束的低秩表示子空间聚类算法

范莉莉, 卢桂馥(), 唐肝翌, 杨丹   

  1. 安徽工程大学 计算机与信息学院,安徽 芜湖 241000
  • 收稿日期:2021-07-08 修回日期:2021-09-03 接受日期:2021-09-06 发布日期:2021-09-16 出版日期:2022-01-10
  • 通讯作者: 卢桂馥
  • 作者简介:范莉莉(1982—),女,山东莱芜人,讲师,硕士,CCF会员,主要研究方向:机器学习、模式识别
    卢桂馥(1976—),男,浙江东阳人,教授,博士,主要研究方向:人工智能、模式识别
    唐肝翌(1977—),男,江西安远人,副教授,硕士,CCF会员,主要研究方向:机器学习、图像处理
    杨丹(1973—),男,湖南长沙人,副教授,硕士,主要研究方向:图像处理、视频编码。
  • 基金资助:
    国家自然科学基金资助项目(61976005);安徽省高等教育提升计划省级自然科学研究一般项目(TSKJ2016B01)

Low-rank representation subspace clustering method based on Hessian regularization and non-negative constraint

Lili FAN, Guifu LU(), Ganyi TANG, Dan YANG   

  1. School of Computer and Information,Anhui Polytechnic University,Wuhu Anhui 241000,China
  • Received:2021-07-08 Revised:2021-09-03 Accepted:2021-09-06 Online:2021-09-16 Published:2022-01-10
  • Contact: Guifu LU
  • About author:FAN Lili, born in 1982, M. S., lecturer. Her research interests include machine learning, pattern recognition.
    LU Guifu, born in 1976, Ph. D., professor. His research interests include artificial intelligence, pattern recognition.
    TANG Ganyi, born in 1977, M. S., associate professor. His research interests include machine learning, image processing.
    YANG Dan, born in 1973, M. S., associate professor. His research interests include image processing, video encoding.
  • Supported by:
    National Natural Science Foundation of China(61976005);Natural Science Research General Program of Anhui Higher Education Promotion Plan(TSKJ2016B01)

摘要:

针对低秩表示(LRR)子空间聚类算法没有考虑数据局部结构,在学习中可能会造成局部相似信息丢失的问题,提出了一种基于Hessian正则化和非负约束的低秩表示子空间聚类算法(LRR-HN),用来探索数据的全局结构和局部结构。首先,利用Hessian正则化良好的推测能力来保持数据的局部流形结构,使数据局部拓扑结构的表达能力更强;其次,考虑到获得的系数矩阵往往有正有负,而负值往往没有实际意义的特点,引入非负约束来保证模型解的有效性,使其在数据局部结构描述上更有意义;最后,通过最小化核范数寻求数据全局结构的低秩表示,从而更好地聚类高维数据。此外,利用自适应惩罚的线性交替方向法设计了一种求解LRR-HN的有效算法,并在一些真实数据集上,采用正确率(AC)和归一化互信息(NMI)对所提出的算法进行了评估。在ORL数据集上聚类数目为20时的实验中,LRR-HN与LRR算法相比,AC和NMI分别提高了11%和9.74%;与自适应低秩表示(ALRR)算法相比,AC和NMI分别提高了5%和1.05%。实验结果表明,LRR-HN与现有的一些算法相比,AC和NMI均有较大的提升,有较好的聚类性能。

关键词: 子空间聚类, Hessian正则化, 非负约束, 低秩表示, 流形学习

Abstract:

Focusing on the issue that the Low-Rank Representation (LRR) subspace clustering algorithm does not consider the local structure of the data and may cause the loss of local similar information during learning, a Low-Rank Representation subspace clustering algorithm based on Hessian regularization and Non-negative constraint (LRR-HN) was proposed to explore the global and local structure of the data. Firstly, the good speculative ability of Hessian regularization was used to maintain the local manifold structure of the data, so that the local topological structure of the data was more expressive. Secondly, considering that the obtained coefficient matrix often has positive and negative values, and the negative values often have no practical significance, non-negative constraints were introduced to ensure the effectiveness of the model solution and make it more meaningful in the description of the local structure of the data. Finally, the low-rank representation of the global structure of the data was sought by minimizing the nuclear norm, so as to cluster high-dimensional data better. In addition, an effective algorithm for solving LRR-HN was designed by using the linearized alternating direction method with adaptive penalty, and the proposed algorithm was evaluated by ACcuracy (AC) and Normalized Mutual Information (NMI) on some real datasets. In the experiments with clusters number 20 on ORL dataset, compared with LRR algorithm, LRR-HN has the AC and NMI increased by 11% and 9.74% respectively, and compared with Adaptive Low-Rank Representation (ALRR) algorithm, LRR-HN has the AC and NMI increased by 5% and 1.05% respectively. Experimental results show that the LRR-HN has great improvement in AC and NMI compared with some existing algorithms, and has the excellent clustering performance.

Key words: subspace clustering, Hessian regularization, non-negative constraint, low-rank representation, manifold learning

中图分类号: