Showing 1–3 of 3 results for author: Wright, O

Edge Entropy as an Indicator of the Effectiveness of GNNs over CNNs for Node Classification

Authors: Lavender Yao Jiang, John Shi, Mark Cheung, Oren Wright, José M. F. Moura

Abstract: Graph neural networks (GNNs) extend convolutional neural networks (CNNs) to graph-based data. A question that arises is how much performance improvement does the underlying graph structure in the GNN provide over the CNN (that ignores this graph structure). To address this question, we introduce edge entropy and evaluate how good an indicator it is for possible performance improvement of GNNs over… ▽ More Graph neural networks (GNNs) extend convolutional neural networks (CNNs) to graph-based data. A question that arises is how much performance improvement does the underlying graph structure in the GNN provide over the CNN (that ignores this graph structure). To address this question, we introduce edge entropy and evaluate how good an indicator it is for possible performance improvement of GNNs over CNNs. Our results on node classification with synthetic and real datasets show that lower values of edge entropy predict larger expected performance gains of GNNs over CNNs, and, conversely, higher edge entropy leads to expected smaller improvement gains. △ Less

Submitted 15 December, 2020; originally announced December 2020.

Graph Signal Processing and Deep Learning: Convolution, Pooling, and Topology

Authors: Mark Cheung, John Shi, Oren Wright, Lavender Y. Jiang, Xujin Liu, José M. F. Moura

Abstract: Deep learning, particularly convolutional neural networks (CNNs), have yielded rapid, significant improvements in computer vision and related domains. But conventional deep learning architectures perform poorly when data have an underlying graph structure, as in social, biological, and many other domains. This paper explores 1)how graph signal processing (GSP) can be used to extend CNN components… ▽ More Deep learning, particularly convolutional neural networks (CNNs), have yielded rapid, significant improvements in computer vision and related domains. But conventional deep learning architectures perform poorly when data have an underlying graph structure, as in social, biological, and many other domains. This paper explores 1)how graph signal processing (GSP) can be used to extend CNN components to graphs in order to improve model performance; and 2)how to design the graph CNN architecture based on the topology or structure of the data graph. △ Less

Submitted 3 August, 2020; originally announced August 2020.

Comments: To be published on IEEE Signal Processing Magazine

Pooling in Graph Convolutional Neural Networks

Authors: Mark Cheung, John Shi, Lavender Yao Jiang, Oren Wright, José M. F. Moura

Abstract: Graph convolutional neural networks (GCNNs) are a powerful extension of deep learning techniques to graph-structured data problems. We empirically evaluate several pooling methods for GCNNs, and combinations of those graph pooling methods with three different architectures: GCN, TAGCN, and GraphSAGE. We confirm that graph pooling, especially DiffPool, improves classification accuracy on popular gr… ▽ More Graph convolutional neural networks (GCNNs) are a powerful extension of deep learning techniques to graph-structured data problems. We empirically evaluate several pooling methods for GCNNs, and combinations of those graph pooling methods with three different architectures: GCN, TAGCN, and GraphSAGE. We confirm that graph pooling, especially DiffPool, improves classification accuracy on popular graph classification datasets and find that, on average, TAGCN achieves comparable or better accuracy than GCN and GraphSAGE, particularly for datasets with larger and sparser graph structures. △ Less

Submitted 7 April, 2020; originally announced April 2020.

Comments: 5 pages, 2 figures, 2019 Asilomar Conference paper