Deep Convolutional Networks on Graph-Structured Data
read the original abstract
Deep Learning's recent successes have mostly relied on Convolutional Networks, which exploit fundamental statistical properties of images, sounds and video data: the local stationarity and multi-scale compositional structure, that allows expressing long range interactions in terms of shorter, localized interactions. However, there exist other important examples, such as text documents or bioinformatic data, that may lack some or all of these strong statistical regularities. In this paper we consider the general question of how to construct deep architectures with small learning complexity on general non-Euclidean domains, which are typically unknown and need to be estimated from the data. In particular, we develop an extension of Spectral Networks which incorporates a Graph Estimation procedure, that we test on large-scale classification problems, matching or improving over Dropout Networks with far less parameters to estimate.
This paper has not been read by Pith yet.
Forward citations
Cited by 8 Pith papers
-
Graph Attention Networks
Graph Attention Networks compute learnable attention coefficients over node neighborhoods to produce weighted feature aggregations, achieving state-of-the-art results on citation networks and inductive protein-protein...
-
Mitigating Structural Overfitting: A Distribution-Aware Rectification Framework for Missing Feature Imputation
DART mitigates structural overfitting in graph missing-feature imputation via global structural augmentation, masked-autoencoder semantic rectification, and test-time distribution rectification, outperforming prior me...
-
How Wide and How Deep? Mitigating Over-Squashing of GNNs via Channel Capacity Constrained Estimation
C3E estimates hidden dimensions and depths for GNNs by treating them as communication channels to reduce over-squashing and improve representation learning.
-
HONEM: Learning Embedding for Higher Order Networks
HONEM learns embeddings for higher-order networks capturing non-Markovian dependencies and outperforms baselines on node classification, reconstruction, link prediction, and visualization.
-
Spectral-based Graph Convolutional Network for Directed Graphs
A spectral-based GCN for directed graphs uses redefined Laplacians to enable direct application to directed data and outperforms prior methods on semi-supervised node classification tasks.
-
Image Classification with Hierarchical Multigraph Networks
Hierarchical multigraph GCNs applied to superpixels achieve competitive or superior accuracy to CNNs on standard image classification benchmarks.
-
Motorway Traffic Flow Prediction using Advanced Deep Learning
Deep learning architectures (CNN, RNN, CNN-LSTM) are applied to motorway traffic data and shown to outperform traditional methods for multi-station, multi-horizon flow prediction.
-
Low-Rank Matrix Completion: A Contemporary Survey
A survey classifying low-rank matrix completion techniques into two categories, discussing required matrix properties, CNN-based variants, and comparing recovery performance with computational complexity.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.