Multi-view Graph Convolutional Networks with Differentiable Node Selection

Zhaoliang Chen,Lele Fu,Shunxin Xiao,Shiping Wang,Claudia Plant,Wenzhong Guo

DOI: https://doi.org/10.1145/3608954

2023-08-13

Abstract:Multi-view data containing complementary and consensus information can facilitate representation learning by exploiting the intact integration of multi-view features. Because most objects in real world often have underlying connections, organizing multi-view data as heterogeneous graphs is beneficial to extracting latent information among different objects. Due to the powerful capability to gather information of neighborhood nodes, in this paper, we apply Graph Convolutional Network (GCN) to cope with heterogeneous-graph data originating from multi-view data, which is still under-explored in the field of GCN. In order to improve the quality of network topology and alleviate the interference of noises yielded by graph fusion, some methods undertake sorting operations before the graph convolution procedure. These GCN-based methods generally sort and select the most confident neighborhood nodes for each vertex, such as picking the top-k nodes according to pre-defined confidence values. Nonetheless, this is problematic due to the non-differentiable sorting operators and inflexible graph embedding learning, which may result in blocked gradient computations and undesired performance. To cope with these issues, we propose a joint framework dubbed Multi-view Graph Convolutional Network with Differentiable Node Selection (MGCN-DNS), which is constituted of an adaptive graph fusion layer, a graph learning module and a differentiable node selection schema. MGCN-DNS accepts multi-channel graph-structural data as inputs and aims to learn more robust graph fusion through a differentiable neural network. The effectiveness of the proposed method is verified by rigorous comparisons with considerable state-of-the-art approaches in terms of multi-view semi-supervised classification tasks.

Machine Learning

What problem does this paper attempt to address?

The paper aims to address several key issues in semi-supervised classification tasks with multi-view data: 1. **Multi-view Data Fusion**: The paper proposes a new framework—Multi-view Graph Convolutional Network with Differentiable Node Selection (MGCN-DNS) to handle multi-view data and integrate complementary information from different views through a graph fusion layer. 2. **Noise Interference**: Existing methods perform sorting operations before graph convolution to select the most relevant neighbor nodes for each vertex, but these operations are often non-differentiable, leading to obstructed gradient propagation. MGCN-DNS proposes a differentiable node selection process to solve this problem. 3. **Multi-view Extension of Graph Convolutional Networks**: Traditional Graph Convolutional Networks (GCNs) are only suitable for single-view data. This paper extends GCNs to handle multi-view structured data through a two-stage adaptive graph fusion process. In summary, the main goal of this paper is to develop a method that can effectively handle multi-view data and achieve better performance in semi-supervised classification tasks. By introducing a differentiable node selection mechanism, MGCN-DNS can automatically select important nodes and maintain the differentiability of the neural network, thereby improving the overall performance of the model.

Multi-view Graph Convolutional Networks with Differentiable Node Selection

Learnable Graph Convolutional Network and Feature Fusion for Multi-view Learning

Adaptive Multi-Channel Contrastive Graph Convolutional Network with Graph and Feature Fusion

Heterogeneous graph convolutional network for multi-view semi-supervised classification

Attention-based stackable graph convolutional network for multi-view learning

Revisiting multi-view learning: A perspective of implicitly heterogeneous Graph Convolutional Network

Semi-Supervised and Self-Supervised Classification with Multi-View Graph Neural Networks

Multi-view Heterogeneous Graph Neural Networks for Node Classification

Interpretable Graph Convolutional Network for Multi-View Semi-Supervised Learning

Multi-View Semi-Supervised Feature Selection with Graph Convolutional Networks

Multi-view Contrastive Learning for Multilayer Network Embedding

Co-GCN for Multi-View Semi-Supervised Learning.

Multi-Channel Graph Convolutional Networks for Graphs with Inconsistent Structures and Features

MVMA-GCN: Multi-view Multi-Layer Attention Graph Convolutional Networks

Multi-view Dual-Channel Graph Convolutional Networks with Multi-Task Learning

Geometric Localized Graph Convolutional Network for Multi-view Semi-supervised Classification

An Efficient Multi-View Heterogeneous Hypergraph Convolutional Network for Heterogeneous Information Network Representation Learning

Multi-view Enhanced Zero-Shot Node Classification

Graph Convolutional Network With Self-Augmented Weights for Semi-Supervised Multi-View Learning

DeepMCGCN: Multi-channel Deep Graph Neural Networks

AM-GCN: Adaptive Multi-channel Graph Convolutional Networks