Summary of Revisiting Edge Perturbation For Graph Neural Network in Graph Data Augmentation and Attack, by Xin Liu et al.
Revisiting Edge Perturbation for Graph Neural Network in Graph Data Augmentation and Attack
by Xin Liu, Yuxiang Zhang, Meng Wu, Mingyu Yan, Kun He, Wei Yan, Shirui Pan, Xiaochun Ye, Dongrui Fan
First submitted to arxiv on: 10 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Cryptography and Security (cs.CR)
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 The abstract presents research on edge perturbation methods for modifying graph structures, which can be categorized into two veins: graph data augmentation and attack. These methods share the same operations but have opposite effects on the accuracy of graph neural networks (GNNs). The study highlights the need to define a clear boundary between these methods, as inappropriate perturbations can lead to undesirable outcomes. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Edge perturbation is used to modify graph structures. There are two types: one helps GNNs learn better, and the other hurts their performance. These methods use the same techniques but get different results. The study wants to figure out why this happens and how to make edge perturbation work well. |
Keywords
* Artificial intelligence * Data augmentation
