Summary of Variational Graph Auto-encoder Based Inductive Learning Method For Semi-supervised Classification, by Hanxuan Yang et al.
Variational Graph Auto-Encoder Based Inductive Learning Method for Semi-Supervised Classification
by Hanxuan Yang, Zhaoxin Yu, Qingchao Kong, Wei Liu, Wenji Mao
First submitted to arxiv on: 26 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 The paper proposes a novel framework for graph representation learning, specifically focusing on variational graph auto-encoders (VGAEs) for inductive learning tasks. Unlike traditional GNN-based methods that heavily rely on annotated nodes, VGAEs are more generalizable and have achieved promising results on unsupervised learning tasks. However, there is still a lack of work on leveraging the VGAE framework for inductive learning due to difficulties in training and avoiding over-fitting. To address these issues, the authors propose Self-Label Augmented VGAE (SLAVG) model that incorporates node labels as one-hot encoded inputs for label reconstruction during training. Additionally, they introduce the Self-Label Augmentation Method (SLAM) to enhance label information using pseudo labels generated by the model with a node-wise masking approach. The proposed model is evaluated on benchmark inductive learning graph datasets, demonstrating superior performance under semi-supervised learning settings. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper explores ways to improve graph representation learning using variational graph auto-encoders (VGAEs). Traditional methods rely too heavily on labeled data, but VGAEs can capture internal structure without labels. However, there’s a gap in training these models for new, unseen structures during inference. To bridge this gap, the authors propose a new model that uses node labels to train and then generates pseudo-labels to help with semi-supervised learning. This approach outperforms others on benchmark datasets. |
Keywords
* Artificial intelligence * Gnn * Inference * One hot * Representation learning * Semi supervised * Unsupervised
