Summary of Ses: Bridging the Gap Between Explainability and Prediction Of Graph Neural Networks, by Zhenhua Huang et al.
SES: Bridging the Gap Between Explainability and Prediction of Graph Neural Networks
by Zhenhua Huang, Kunhao Li, Shaojie Wang, Zhaohong Jia, Wentao Zhu, Sharad Mehrotra
First submitted to arxiv on: 16 Jul 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 addresses the challenges of achieving high-accuracy and interpretable predictions with Graph Neural Networks (GNNs). Existing GNN interpreters provide post-hoc explanations that can be misrepresentative, while self-explained GNNs offer built-in explanations during training but struggle to generate high-quality node feature and subgraph explanations. The authors propose a Self-Explainable and Self-Supervised graph neural network (SES) to bridge this gap. SES consists of two processes: explainable training and enhanced predictive learning. During explainable training, SES generates crucial structure and feature masks using a global mask generator co-trained with a graph encoder, reducing time consumption and providing node feature and subgraph explanations. In the enhanced predictive learning phase, mask-based positive-negative pairs are constructed to compute a triplet loss and enhance node representations through contrastive learning. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about making computer programs that understand graph data better. Right now, these programs can’t always explain why they made certain predictions. The authors want to fix this problem by creating a new kind of program called Self-Explainable and Self-Supervised (SES). This SES program has two parts: one part helps the program learn what’s important in the graph data, and another part uses that information to make better predictions. |
Keywords
* Artificial intelligence * Encoder * Gnn * Graph neural network * Mask * Self supervised * Triplet loss
