A Simple Way to Learn Metrics Between Attributed Graphs

Yacouba Kaloga, Pierre Borgnat, Amaury Habrard
Proceedings of the First Learning on Graphs Conference, PMLR 198:25:1-25:12, 2022.

Abstract

The choice of good distances and similarity measures between objects is important for many machine learning methods. Therefore, many metric learning algorithms have been developed in recent years, mainly for Euclidean data in order to improve performance of classification or clustering methods. However, due to difficulties in establishing computable, efficient and differentiable distances between attributed graphs, few metric learning algorithms adapted to graphs have been developed despite the strong interest of the community. In this paper, we address this issue by proposing a new Simple Graph Metric Learning - SGML - model with few trainable parameters based on Simple Convolutional Neural Networks - SGCN - and elements of optimal transport theory. This model allows us to build an appropriate distance from a database of labeled (attributed) graphs to improve the performance of simple classification algorithms such as \textdollar k\textdollar -NN. This distance can be quickly trained while maintaining good performances as illustrated by the experimental study presented in this paper.

Cite this Paper

BibTeX
@InProceedings{pmlr-v198-kaloga22a, title = {A Simple Way to Learn Metrics Between Attributed Graphs}, author = {Kaloga, Yacouba and Borgnat, Pierre and Habrard, Amaury}, booktitle = {Proceedings of the First Learning on Graphs Conference}, pages = {25:1--25:12}, 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/kaloga22a/kaloga22a.pdf}, url = {https://proceedings.mlr.press/v198/kaloga22a.html}, abstract = {The choice of good distances and similarity measures between objects is important for many machine learning methods. Therefore, many metric learning algorithms have been developed in recent years, mainly for Euclidean data in order to improve performance of classification or clustering methods. However, due to difficulties in establishing computable, efficient and differentiable distances between attributed graphs, few metric learning algorithms adapted to graphs have been developed despite the strong interest of the community. In this paper, we address this issue by proposing a new Simple Graph Metric Learning - SGML - model with few trainable parameters based on Simple Convolutional Neural Networks - SGCN - and elements of optimal transport theory. This model allows us to build an appropriate distance from a database of labeled (attributed) graphs to improve the performance of simple classification algorithms such as \textdollar k\textdollar -NN. This distance can be quickly trained while maintaining good performances as illustrated by the experimental study presented in this paper.} }
Endnote
%0 Conference Paper %T A Simple Way to Learn Metrics Between Attributed Graphs %A Yacouba Kaloga %A Pierre Borgnat %A Amaury Habrard %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-kaloga22a %I PMLR %P 25:1--25:12 %U https://proceedings.mlr.press/v198/kaloga22a.html %V 198 %X The choice of good distances and similarity measures between objects is important for many machine learning methods. Therefore, many metric learning algorithms have been developed in recent years, mainly for Euclidean data in order to improve performance of classification or clustering methods. However, due to difficulties in establishing computable, efficient and differentiable distances between attributed graphs, few metric learning algorithms adapted to graphs have been developed despite the strong interest of the community. In this paper, we address this issue by proposing a new Simple Graph Metric Learning - SGML - model with few trainable parameters based on Simple Convolutional Neural Networks - SGCN - and elements of optimal transport theory. This model allows us to build an appropriate distance from a database of labeled (attributed) graphs to improve the performance of simple classification algorithms such as \textdollar k\textdollar -NN. This distance can be quickly trained while maintaining good performances as illustrated by the experimental study presented in this paper.
APA
Kaloga, Y., Borgnat, P. & Habrard, A.. (2022). A Simple Way to Learn Metrics Between Attributed Graphs. Proceedings of the First Learning on Graphs Conference, in Proceedings of Machine Learning Research 198:25:1-25:12 Available from https://proceedings.mlr.press/v198/kaloga22a.html.

Related Material