Learning to Predict Graphs with Fused Gromov-Wasserstein Barycenters
This addresses graph prediction tasks, particularly for applications like metabolic identification, with a novel optimal transport-based approach.
The paper tackles the problem of supervised labeled graph prediction by formulating it as regression with a Fused Gromov-Wasserstein loss and proposing models based on FGW barycenters. The method achieves very good performance on a metabolic identification problem with minimal engineering.
This paper introduces a novel and generic framework to solve the flagship task of supervised labeled graph prediction by leveraging Optimal Transport tools. We formulate the problem as regression with the Fused Gromov-Wasserstein (FGW) loss and propose a predictive model relying on a FGW barycenter whose weights depend on inputs. First we introduce a non-parametric estimator based on kernel ridge regression for which theoretical results such as consistency and excess risk bound are proved. Next we propose an interpretable parametric model where the barycenter weights are modeled with a neural network and the graphs on which the FGW barycenter is calculated are additionally learned. Numerical experiments show the strength of the method and its ability to interpolate in the labeled graph space on simulated data and on a difficult metabolic identification problem where it can reach very good performance with very little engineering.