Covered Forest: Fine-grained generalization analysis of graph neural networks
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912149952004096 |
|---|---|
| author | Vasileiou, Antonis Finkelshtein, Ben Geerts, Floris Levie, Ron Morris, Christopher |
| author_facet | Vasileiou, Antonis Finkelshtein, Ben Geerts, Floris Levie, Ron Morris, Christopher |
| contents | The expressive power of message-passing graph neural networks (MPNNs) is reasonably well understood, primarily through combinatorial techniques from graph isomorphism testing. However, MPNNs' generalization abilities -- making meaningful predictions beyond the training set -- remain less explored. Current generalization analyses often overlook graph structure, limit the focus to specific aggregation functions, and assume the impractical, hard-to-optimize $0$-$1$ loss function. Here, we extend recent advances in graph similarity theory to assess the influence of graph structure, aggregation, and loss functions on MPNNs' generalization abilities. Our empirical study supports our theoretical insights, improving our understanding of MPNNs' generalization properties. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_07106 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Covered Forest: Fine-grained generalization analysis of graph neural networks Vasileiou, Antonis Finkelshtein, Ben Geerts, Floris Levie, Ron Morris, Christopher Machine Learning Discrete Mathematics Data Structures and Algorithms Neural and Evolutionary Computing The expressive power of message-passing graph neural networks (MPNNs) is reasonably well understood, primarily through combinatorial techniques from graph isomorphism testing. However, MPNNs' generalization abilities -- making meaningful predictions beyond the training set -- remain less explored. Current generalization analyses often overlook graph structure, limit the focus to specific aggregation functions, and assume the impractical, hard-to-optimize $0$-$1$ loss function. Here, we extend recent advances in graph similarity theory to assess the influence of graph structure, aggregation, and loss functions on MPNNs' generalization abilities. Our empirical study supports our theoretical insights, improving our understanding of MPNNs' generalization properties. |
| title | Covered Forest: Fine-grained generalization analysis of graph neural networks |
| topic | Machine Learning Discrete Mathematics Data Structures and Algorithms Neural and Evolutionary Computing |
| url | https://arxiv.org/abs/2412.07106 |