Summary of Navigating Complexity: Toward Lossless Graph Condensation Via Expanding Window Matching, by Yuchen Zhang and Tianle Zhang and Kai Wang and Ziyao Guo and Yuxuan Liang and Xavier Bresson and Wei Jin and Yang You
Navigating Complexity: Toward Lossless Graph Condensation via Expanding Window Matching
by Yuchen Zhang, Tianle Zhang, Kai Wang, Ziyao Guo, Yuxuan Liang, Xavier Bresson, Wei Jin, Yang You
First submitted to arxiv on: 7 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
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 This paper tackles the challenge of condensing large-scale graph datasets while preserving the performance of Graph Neural Networks (GNNs) trained on them. Existing methods often fail to accurately replicate the original graph, leading to biased and restricted supervision signals that limit the scale and efficacy of the condensed graph. The authors make a first attempt at lossless graph condensation by bridging this gap through curriculum learning and expanding window matching. They design a loss function to extract knowledge from expert trajectories and demonstrate the superiority of their method across different datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about making big graphs smaller without losing any important information. Right now, there are ways to shrink graphs, but they don’t always work well. The authors found out why this happens and then created a new way to make sure that the shrunk graph has all the same good qualities as the original one. They tested their method on many different graphs and showed that it works better than other methods. |
Keywords
* Artificial intelligence * Curriculum learning * Loss function
