Incomplete Multiview Clustering via Cross-View Relation Transfer
Incomplete Multiview Clustering via Cross-View Relation Transfer
复制标题
基于跨视图关系转移的不完全多视图聚类
DOI：
10.1109/tcsvt.2022.3201822
复制
发表时间：
2021-12
影响因子：
8.4
通讯作者：
Yiming Wang;Dongxia Chang;Zhiqiang Fu;Jie Wen;Yao Zhao
中科院分区：
文献类型：
--
作者：
Yiming Wang;Dongxia Chang;Zhiqiang Fu;Jie Wen;Yao Zhao
本文研究了不完全视图上的多视图聚类问题。与完全多视图聚类相比，视图缺失问题增加了从不同视图中学习共同表示的难度。为了解决这一挑战，我们提出了一种新的不完全多视图聚类框架，该框架结合了跨视图关系迁移和多视图融合学习。具体而言，基于多视图数据中存在的一致性，设计了基于跨视图关系迁移的补全模块，该模块将已知的相似实例间关系迁移到缺失视图中，并基于迁移的关系图通过图网络对缺失数据进行推断。然后设计了特定于视图的编码器来提取恢复的多视图数据，并引入了基于注意力的融合层来获得公共表示。此外，为了减少视图不一致带来的误差影响，获得更好的聚类结构，引入联合聚类层对恢复和聚类同时进行优化。在多个真实数据集上进行的大量实验证明了该方法的有效性。
In this paper, we consider the problem of multi-view clustering on incomplete views. Compared with complete multi-view clustering, the view-missing problem increases the difficulty of learning common representations from different views. To address the challenge, we propose a novel incomplete multi-view clustering framework, which incorporates cross-view relation transfer and multi-view fusion learning. Specifically, based on the consistency existing in multi-view data, we devise a cross-view relation transfer-based completion module, which transfers known similar inter-instance relationships to the missing view and infers the missing data via graph networks based on the transferred relationship graph. Then the view-specific encoders are designed to extract the recovered multi-view data, and an attention-based fusion layer is introduced to obtain the common representation. Moreover, to reduce the impact of the error caused by the inconsistency between views and obtain a better clustering structure, a joint clustering layer is introduced to optimize recovery and clustering simultaneously. Extensive experiments conducted on several real datasets demonstrate the effectiveness of the proposed method.