Summary of A Teacher-free Graph Knowledge Distillation Framework with Dual Self-distillation, by Lirong Wu et al.
A Teacher-Free Graph Knowledge Distillation Framework with Dual Self-Distillation
by Lirong Wu, Haitao Lin, Zhangyang Gao, Guojiang Zhao, Stan Z. Li
First submitted to arxiv on: 6 Mar 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 In this paper, the authors propose a novel approach to graph knowledge distillation called Teacher-Free Graph Self-Distillation (TGS), which eliminates the need for teacher models or Graph Neural Networks (GNNs). Instead, TGS uses Multi-Layer Perceptrons (MLPs) and relies on structural information to guide dual self-distillation between target nodes and their neighborhoods. The authors demonstrate that vanilla MLPs can be significantly improved with this approach, achieving an average improvement of 15.54% over existing methods on six real-world datasets. Additionally, TGS infers 75X-89X faster than existing GNNs and 16X-25X faster than classical inference acceleration methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Graph knowledge distillation is a technique that helps improve the performance of smaller models by transferring knowledge from larger, more experienced models. In this case, the authors use Multi-Layer Perceptrons (MLPs) instead of Graph Neural Networks (GNNs), which are typically used for graph-related tasks. The proposed approach, called Teacher-Free Graph Self-Distillation (TGS), uses structural information to guide self-distillation between nodes and their neighborhoods. This allows the model to learn from its own mistakes and improve its performance without needing a teacher model or GNN. |
Keywords
* Artificial intelligence * Distillation * Gnn * Inference * Knowledge distillation * Teacher model
