Summary of Improving the Interpretability Of Gnn Predictions Through Conformal-based Graph Sparsification, by Pablo Sanchez-martin et al.
Improving the interpretability of GNN predictions through conformal-based graph sparsification
by Pablo Sanchez-Martin, Kinaan Aamir Khan, Isabel Valera
First submitted to arxiv on: 18 Apr 2024
Categories
- Main: Machine Learning (stat.ML)
- Secondary: Machine Learning (cs.LG); Social and Information Networks (cs.SI)
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 A novel approach to training Graph Neural Networks (GNNs) is proposed, which jointly identifies the most predictive subgraph and optimizes graph classification task performance. This method, utilizing reinforcement learning and conformal predictions, enables GNNs to rely on significantly sparser subgraphs while matching state-of-the-art performance. The approach is evaluated on nine graph classification datasets, demonstrating competitive results and improved interpretability. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary A new way of training Graph Neural Networks (GNNs) is being explored. Instead of using all the information in a graph, this method finds the most important parts that help with predictions. It does this by removing edges and nodes without knowing what the subgraph should look like beforehand. The approach uses something called reinforcement learning to make decisions and also considers how unsure the GNN is about its predictions. Tests on nine different types of graphs show that this method works well and provides more understandable results. |
Keywords
» Artificial intelligence » Classification » Gnn » Reinforcement learning
