Summary of Unifying Invariance and Spuriousity For Graph Out-of-distribution Via Probability Of Necessity and Sufficiency, by Xuexin Chen et al.
Unifying Invariance and Spuriousity for Graph Out-of-Distribution via Probability of Necessity and Sufficiency
by Xuexin Chen, Ruichu Cai, Kaitao Zheng, Zhifan Jiang, Zhengting Huang, Zhifeng Hao, Zijian Li
First submitted to arxiv on: 14 Feb 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 This paper proposes a unified framework to address the challenge of generalizing models trained on biased data to unseen test data. The framework, called PNSIS, leverages the Probability of Necessity and Sufficiency (PNS) to extract the invariant substructure from graph data. This is achieved by minimizing an upper bound built on theoretical advances in PNS. Additionally, PNSIS involves invariant and spurious subgraph classifiers to enhance generalization performance. Experimental results show that PNSIS outperforms state-of-the-art techniques on several benchmarks, demonstrating its effectiveness in real-world scenarios. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper tackles the important problem of graph out-of-distribution (OOD) by developing a new framework called PNSIS. This framework helps models trained on biased data to generalize better to unseen test data. The authors use special math ideas to find the most important parts of the graph that don’t change when the environment changes. They also add special classifiers to help the model make good predictions even with noisy data. The results show that PNSIS works well and beats other methods on several tests. |
Keywords
* Artificial intelligence * Generalization * Probability
