Summary of Coefficient Decomposition For Spectral Graph Convolution, by Feng Huang et al.
Coefficient Decomposition for Spectral Graph Convolution
by Feng Huang, Wen Zhang
First submitted to arxiv on: 6 May 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
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 proposes a general form for spectral graph convolutional networks (SGCNs) based on graph signal filters, which has shown promise in modeling graph-structured data. The authors explore various polynomial basis coefficients and model architectures, showing that existing SGCNs can be derived from the proposed framework by coefficient decomposition operations. Building on this foundation, the paper introduces two novel spectral graph convolutions: CoDeSGC-CP and -Tucker, which apply tensor decomposition techniques to improve performance. Experimental results demonstrate significant improvements in performance. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Spectral graph convolutional networks (SGCNs) are a type of graph neural network that uses filters to model graph-structured data. The authors of this paper explore ways to improve SGCNs by looking at the coefficients used in these filters and how they’re arranged. They show that existing SGCNs can be built using their new approach, and use this idea to create two new types of SGNs that do better than before. This is important because it could help with things like social network analysis and recommendation systems. |
Keywords
» Artificial intelligence » Graph neural network
