Research map / Graph learning / GNN architectures
Digraph Inception Convolutional Networks
Zekun Tong, Yuxuan Liang, Changsheng Sun, Xinke Li, David Rosenblum, Andrew Lim
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Inception block (k-th order proximity) to digraph convolution (ablation) | DiGCN w/ Lappr · CORA-ML | accuracy (%): 77.01 -> 80.28 | Improved |
| add Inception block (k-th order proximity) to digraph convolution (ablation) | DiGCN w/ Lappr · CITESEER | accuracy (%): 64.92 -> 66.11 | Improved |
| add Inception block (k-th order proximity) to digraph convolution (ablation) | DiGCN w/ Lappr · AM-PHOTO | accuracy (%): 88.72 -> 90.02 | Improved |
| add Inception block (k-th order proximity) to digraph convolution (ablation) | DiGCN w/ Lappr · AM-COMPUTER | accuracy (%): 85.55 -> 85.94 | Improved |
| use sparse personalized-PageRank Laplacian Lappr instead of dense PageRank Laplacian Lpr (ablation) | DiGCN w/o IB · CORA-ML | accuracy (%): 77.11 -> 77.01 | No clear effect |
| use sparse personalized-PageRank Laplacian Lappr instead of dense Lpr (ablation) | DiGCN w/o IB · AM-PHOTO | accuracy (%): OOM -> 88.72 | Improved |
| use directed PageRank digraph convolution instead of undirected GCN | GCN · CORA-ML | accuracy (%): 53.11 -> 77.01 | Improved |
| increase Inception layers from 1 to 3 (ablation) | DiGCN · CORA-ML | accuracy (%): 72.14 -> 80.28 | Improved |
| increase Inception layers from 3 to 4 (ablation) | DiGCN · CORA-ML | accuracy (%): 80.28 -> 79.95 | No clear effect |
| increase model width k beyond 2 (ablation) | DiGCN single layer · CITESEER | accuracy (%) | No clear effect |
| sum vs concatenation fusion (ablation) | DiGCN · AM-PHOTO | accuracy (%) | Mixed |
| use ReLU non-linear activation in Inception block (ablation) | DiGCN · AM-PHOTO | accuracy (%) | Hurt |
Closest papers by meaning
- Simplifying Graph Convolutional Networks (ICML 2019 · similarity 0.90)
- Break the Ceiling: Stronger Multi-scale Deep Graph Convolutional Networks (NeurIPS 2019 · similarity 0.90)
- Deeper Insights Into Graph Convolutional Networks for Semi-Supervised Learning (AAAI 2018 · similarity 0.90)
- Simple and Deep Graph Convolutional Networks (ICML 2020 · similarity 0.89)
- Learning Discrete Structures for Graph Neural Networks (ICML 2019 · similarity 0.89)
- Graph Learning in 4D: A Quaternion-Valued Laplacian to Enhance Spectral GCNs (AAAI 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
