计算机应用 ›› 2017, Vol. 37 ›› Issue (10): 2834-2840.DOI: 10.11772/j.issn.1001-9081.2017.10.2834

• 人工智能 • 上一篇    下一篇

基于特征融合的多约束非负矩阵分解算法

孙静, 蔡希彪, 孙福明   

  1. 辽宁工业大学 电子与信息工程学院, 辽宁 锦州 121001
  • 收稿日期:2017-02-24 修回日期:2017-04-19 出版日期:2017-10-10 发布日期:2017-10-16
  • 通讯作者: 孙静(1992-),女,辽宁阜新人,硕士研究生,主要研究方向:图像语义理解,E-mail:sunjing616@foxmail.com
  • 作者简介:孙静(1992-),女,辽宁阜新人,硕士研究生,主要研究方向:图像语义理解;蔡希彪(1972-),男,辽宁盘锦人,副教授,博士,主要研究方向:图像语义理解、移动通信;孙福明(1972-),男,辽宁大连人,教授,博士,CCF会员,主要研究方向:图像语义理解、机器学习.
  • 基金资助:
    国家自然科学基金资助项目(61572214);辽宁省高等学校优秀人才支持计划项目(LR2015030)。

Multi-constraint nonnegative matrix factorization algorithm based on feature fusion

SUN Jing, CAI Xibiao, SUN Fuming   

  1. School of Electronics and Information Engineering, Liaoning University of Technology, Jinzhou Liaoning 121001, China
  • Received:2017-02-24 Revised:2017-04-19 Online:2017-10-10 Published:2017-10-16
  • Supported by:
    This work is partially supported by the National Natural Science Foundation of China (61572214), the Program for Liaoning Excelent Talents in University (LR2015030).

摘要: 针对非负矩阵分解后数据的稀疏性降低、单一图像特征不能够很好地描述图像内容的问题,提出一种基于特征融合的多约束非负矩阵分解算法。该算法不仅考虑了少量已知样本的标签信息和稀疏约束,还对其进行了图正则化处理,而且将分解后的具有不同稀疏度的图像特征进行了融合,从而增强了算法的聚类性能和有效性。在Yale-32和COIL20数据集上进行的对比实验进一步验证了该算法具有更好的聚类精度和稀疏性。

关键词: 非负矩阵分解, 标签信息, 稀疏约束, 图正则, 特征融合

Abstract: Focusing on the issues that the sparseness of data is reduced after factorization and the single image feature cannot describe the image content well, a multi-constraint nonnegative matrix factorization based on feature fusion was proposed. The information provided by few known labeled samples and sparseness constraint were considered, and the graph regularization was processed, then the decomposed image features with different sparseness were fused, which improved the clustering performance and effectiveness. Extensive experiments were conducted on both Yale-32 and COIL20 datasets, and the comparisons with four state-of-the-art algorithms demonstrate that the proposed method has superiority in both clustering accuracy and sparseness.

Key words: Non-negative Matrix Factorization (NMF), label information, sparseness constraint, graph regularization, feature fusion

中图分类号: