Performance evaluation of Graph Convolutional Networks with Siamese training for few-shot classification of nodes
Nichita Uțiu
Bridge between perception and reasoning: Graph neural networks and beyond - ICML 2020 workshop (2020) .
2020
Full text: https://logicalreasoninggnn.github.io/
Abstract
Siamese networks learn embeddings which impose geometric constraints on the embedding space and are used in the context of one-few-shot learning. In this paper, we empirically investigate applying this framework to Graph Convolutional Networks (GCNs). We test whether some lightweight architectures yield performance increases over plain Multi-Layer Perceptrons (MLPs) in tasks of one-/few-shot learning for nodes. We show that, for our benchmark, good performance can be achieved even with a fast Simplified Graph Convolutional Network (SGCN).
Notice: Undefined index: publicationsCaching in /www/html/epistemio/application/controllers/PublicationController.php on line 2240
Share comment