Summary of Edge-splitting Mlp: Node Classification on Homophilic and Heterophilic Graphs Without Message Passing, by Matthias Kohn et al.
Edge-Splitting MLP: Node Classification on Homophilic and Heterophilic Graphs without Message Passing
by Matthias Kohn, Marcel Hoffmann, Ansgar Scherp
First submitted to arxiv on: 11 Dec 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Machine Learning (stat.ML)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary A novel approach to message passing neural networks (MPNNs) is proposed, which combines the strengths of MPNNs and graph-based methods for node classification on heterophilic graphs. The model, ES-MLP, incorporates edge splitting mechanisms from Edge Splitting GNN (ES-GNN) into a Graph-MLP framework. This allows ES-MLP to learn separate adjacency matrices based on relevant and irrelevant feature pairs, enabling it to perform competitively with homophilic and heterophilic models on multiple datasets. The proposed approach is robust to edge noise during inference and offers significant speedup compared to traditional MPNNs. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary MPNNs have been successful in node classification on homophilic graphs. A new model called ES-MLP combines ideas from graph-based methods with message passing neural networks. This helps the model work well on heterophilic graphs too. The model is fast and doesn’t need edges during inference. It’s a useful tool for people working with graphs. |
Keywords
* Artificial intelligence * Classification * Gnn * Inference
