UGN is an encoder-decoder that combines graph convolution with 2D convolution on a pairwise node-feature matrix, claiming strong results across six graph tasks on twelve datasets.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Unified Graph Networks (UGN): A Deep Neural Framework for Solving Graph Problems
UGN is an encoder-decoder that combines graph convolution with 2D convolution on a pairwise node-feature matrix, claiming strong results across six graph tasks on twelve datasets.