Summary of Efficient Topology-aware Data Augmentation For High-degree Graph Neural Networks, by Yurui Lai et al.
Efficient Topology-aware Data Augmentation for High-Degree Graph Neural Networks
by Yurui Lai, Xiaoyang Lin, Renchi Yang, Hongtao Wang
First submitted to arxiv on: 8 Jun 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 The proposed framework, TADA, aims to improve the performance of graph neural networks (GNNs) on high-degree graphs (HDGs) by addressing over-smoothing and efficiency issues. This is achieved through two key modules: feature expansion with structure embeddings and topology- and attribute-aware graph sparsification. The first module enhances model capacity by encoding graph structure into high-quality structure embeddings using a highly-efficient sketching method. The second module identifies and reduces redundant/noisy edges from the input graph, alleviating over-smoothing and facilitating faster feature aggregations. This approach improves predictive performance on 8 real HDGs for node classification while maintaining efficient training and inference processes. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary TADA is a new way to make graph neural networks work better on big social networks and other complex graphs. These types of graphs are hard for GNNs to understand because they have many neighbors, which can cause problems when the model tries to learn from them. TADA helps by using extra information about the graph’s structure and attributes to make the model more accurate and efficient. This means that TADA can help GNNs learn better on these complex graphs, making it a useful tool for people working with social networks, transaction graphs, and other HDGs. |
Keywords
» Artificial intelligence » Classification » Inference
