Metric Based Few-Shot Graph Classification

Donato Crisostomi, Simone Antonelli, Valentino Maiorca, Luca Moschella, Riccardo Marin, Emanuele Rodolà
Proceedings of the First Learning on Graphs Conference, PMLR 198:33:1-33:22, 2022.

Abstract

Few-shot graph classification is a novel yet promising emerging research field that still lacks the soundness of well-established research domains. Existing works often consider different benchmarks and evaluation settings, hindering comparison and, therefore, scientific progress. In this work, we start by providing an extensive overview of the possible approaches to solving the task, comparing the current state-of-the-art and baselines via a unified evaluation framework. Our findings show that while graph-tailored approaches have a clear edge on some distributions, easily adapted few-shot learning methods generally perform better. In fact, we show that it is sufficient to equip a simple metric learning baseline with a state-of-the-art graph embedder to obtain the best overall results. We then show that straightforward additions at the latent level lead to substantial improvements by introducing i) a task-conditioned embedding space ii) a MixUp-based data augmentation technique. Finally, we release a highly reusable codebase to foster research in the field, offering modular and extensible implementations of all the relevant techniques.

Cite this Paper

BibTeX
@InProceedings{pmlr-v198-crisostomi22a, title = {Metric Based Few-Shot Graph Classification}, author = {Crisostomi, Donato and Antonelli, Simone and Maiorca, Valentino and Moschella, Luca and Marin, Riccardo and Rodol{\`a}, Emanuele}, booktitle = {Proceedings of the First Learning on Graphs Conference}, pages = {33:1--33:22}, year = {2022}, editor = {Rieck, Bastian and Pascanu, Razvan}, volume = {198}, series = {Proceedings of Machine Learning Research}, month = {09--12 Dec}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v198/crisostomi22a/crisostomi22a.pdf}, url = {https://proceedings.mlr.press/v198/crisostomi22a.html}, abstract = {Few-shot graph classification is a novel yet promising emerging research field that still lacks the soundness of well-established research domains. Existing works often consider different benchmarks and evaluation settings, hindering comparison and, therefore, scientific progress. In this work, we start by providing an extensive overview of the possible approaches to solving the task, comparing the current state-of-the-art and baselines via a unified evaluation framework. Our findings show that while graph-tailored approaches have a clear edge on some distributions, easily adapted few-shot learning methods generally perform better. In fact, we show that it is sufficient to equip a simple metric learning baseline with a state-of-the-art graph embedder to obtain the best overall results. We then show that straightforward additions at the latent level lead to substantial improvements by introducing i) a task-conditioned embedding space ii) a MixUp-based data augmentation technique. Finally, we release a highly reusable codebase to foster research in the field, offering modular and extensible implementations of all the relevant techniques.} }
Endnote
%0 Conference Paper %T Metric Based Few-Shot Graph Classification %A Donato Crisostomi %A Simone Antonelli %A Valentino Maiorca %A Luca Moschella %A Riccardo Marin %A Emanuele Rodolà %B Proceedings of the First Learning on Graphs Conference %C Proceedings of Machine Learning Research %D 2022 %E Bastian Rieck %E Razvan Pascanu %F pmlr-v198-crisostomi22a %I PMLR %P 33:1--33:22 %U https://proceedings.mlr.press/v198/crisostomi22a.html %V 198 %X Few-shot graph classification is a novel yet promising emerging research field that still lacks the soundness of well-established research domains. Existing works often consider different benchmarks and evaluation settings, hindering comparison and, therefore, scientific progress. In this work, we start by providing an extensive overview of the possible approaches to solving the task, comparing the current state-of-the-art and baselines via a unified evaluation framework. Our findings show that while graph-tailored approaches have a clear edge on some distributions, easily adapted few-shot learning methods generally perform better. In fact, we show that it is sufficient to equip a simple metric learning baseline with a state-of-the-art graph embedder to obtain the best overall results. We then show that straightforward additions at the latent level lead to substantial improvements by introducing i) a task-conditioned embedding space ii) a MixUp-based data augmentation technique. Finally, we release a highly reusable codebase to foster research in the field, offering modular and extensible implementations of all the relevant techniques.
APA
Crisostomi, D., Antonelli, S., Maiorca, V., Moschella, L., Marin, R. & Rodolà, E.. (2022). Metric Based Few-Shot Graph Classification. Proceedings of the First Learning on Graphs Conference, in Proceedings of Machine Learning Research 198:33:1-33:22 Available from https://proceedings.mlr.press/v198/crisostomi22a.html.

Related Material