《计算机应用》唯一官方网站 ›› 2023, Vol. 43 ›› Issue (9): 2651-2656.DOI: 10.11772/j.issn.1001-9081.2022091394

• 2022第10届CCF大数据学术会议 •    下一篇

深度融合多视图聚类网络

何子仪, 杨燕(), 张熠玲   

  1. 西南交通大学 计算机与人工智能学院,成都 611756
  • 收稿日期:2022-09-12 修回日期:2022-12-20 接受日期:2022-12-28 发布日期:2023-04-04 出版日期:2023-09-10
  • 通讯作者: 杨燕
  • 作者简介:何子仪(1998—),男,湖北黄冈人,硕士研究生,主要研究方向:深度聚类、多视图聚类
    张熠玲(1994—),女,四川德阳人,博士研究生,主要研究方向:多视图学习、时空数据挖掘。
  • 基金资助:
    国家自然科学基金资助项目(61976247)

Multi-view clustering network with deep fusion

Ziyi HE, Yan YANG(), Yiling ZHANG   

  1. College of Computing and Artificial Intelligence,Southwest Jiaotong University,Chengdu Sichuan 611756,China
  • Received:2022-09-12 Revised:2022-12-20 Accepted:2022-12-28 Online:2023-04-04 Published:2023-09-10
  • Contact: Yan YANG
  • About author:HE Ziyi, born in 1998, M. S. candidate. His research interests include deep clustering, multi-view clustering.
    ZHANG Yiling, born in 1994, Ph. D. candidate. Her research interests include multi-view learning, spatial-temporal data mining.
  • Supported by:
    National Natural Science Foundation of China(61976247)

摘要:

现有的深度多视图聚类方法存在以下缺点:1)在对单一视图进行特征提取时,只考虑了样本的属性信息或结构信息,而没有将二者进行融合,导致提取到的特征不能充分表示原始数据的潜在结构;2)将特征提取与聚类划分为两个独立的过程,没有建立两者间的联系,因此无法利用聚类过程优化特征提取过程。针对以上问题,提出一种深度融合多视图聚类网络(DFMCN)。首先,结合自编码器和图卷积自编码器融合样本的属性信息和结构信息,获取每个视图的嵌入空间;然后,通过加权融合获取融合视图嵌入空间并在此空间中进行聚类,并且在聚类过程中采用双层自监督机制优化特征提取过程。在FM(Fashion-MNIST)、HW(HandWritten numerals)、YTF(YouTube Face)数据集上的实验结果表明:DFMCN的准确率高于所有对比方法;在FM数据集上,DFMCN的准确率比次优的CMSC-DCCA(Cross-Modal Subspace Clustering via Deep Canonical Correlation Analysis)方法提高了1.80个百分点,标准化互信息(NMI)高于除CMSC-DCCA和DMSC(Deep Multimodal Subspace Clustering networks)的所有方法1.26~14.84个百分点。实验结果验证了所提方法的有效性。

关键词: 深度学习, 多视图聚类, 深度聚类, 自表示, 自监督

Abstract:

Current deep multi-view clustering methods have the following shortcomings: 1) When feature extraction is carried out for a single view, only attribute information or structural information of the samples is considered, and these two types of information are not integrated. Thus, the extracted features cannot fully represent latent structure of the original data. 2) Feature extraction and clustering were divided into two separated processes, without establishing the relationship between them, so that the feature extraction process cannot be optimized by the clustering process. To solve these problems, a Deep Fusion based Multi-view Clustering Network (DFMCN) was proposed. Firstly, the embedding space of each view was obtained by combining autoencoder and graph convolution autoencoder to fuse attribute information and structure information of samples. Then, the embedding space of the fusion view was obtained through weighted fusion, and clustering was carried out in this space. And in the process of clustering, the feature extraction process was optimized by a two-layer self-supervision mechanism. Experimental results on FM (Fashion-MNIST), HW (HandWritten numerals), and YTF (YouTube Face) datasets show that the accuracy of DFMCN is higher than those of all comparison methods; and DFMCN has the accuracy increased by 1.80 percentage points compared with the suboptimal CMSC-DCCA (Cross-Modal Subspace Clustering via Deep Canonical Correlation Analysis) method on FM dataset, the Normalized Mutual Information (NMI) of DFMCN is increased by 1.26 to 14.84 percentage points compared to all methods except for CMSC-DCCA and DMSC (Deep Multimodal Subspace Clustering networks). Experimental results verify the effectiveness of the proposed method.

Key words: deep learning, multi-view clustering, deep clustering, self-representation, self-supervision

中图分类号: