LGRPool: Hierarchical Graph Pooling Via Local-Global Regularisation
This addresses a specific limitation in graph neural networks for researchers in graph representation learning, though it appears incremental.
The paper tackles the problem of hierarchical graph pooling methods not aligning local and global features in multiscale graph analysis, proposing LGRPool which uses expectation maximization with regularization to align these aspects. Experimental results show it slightly outperforms baselines on graph classification benchmarks.
Hierarchical graph pooling(HGP) are designed to consider the fact that conventional graph neural networks(GNN) are inherently flat and are also not multiscale. However, most HGP methods suffer not only from lack of considering global topology of the graph and focusing on the feature learning aspect, but also they do not align local and global features since graphs should inherently be analyzed in a multiscale way. LGRPool is proposed in the present paper as a HGP in the framework of expectation maximization in machine learning that aligns local and global aspects of message passing with each other using a regularizer to force the global topological information to be inline with the local message passing at different scales through the representations at different layers of HGP. Experimental results on some graph classification benchmarks show that it slightly outperforms some baselines.