In-n-Out: Calibrating Graph Neural Networks for Link Prediction

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Nascimento, Erik, Mesquita, Diego, Kaski, Samuel, Souza, Amauri H
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866916151815045120
author Nascimento, Erik
Mesquita, Diego
Kaski, Samuel
Souza, Amauri H
author_facet Nascimento, Erik
Mesquita, Diego
Kaski, Samuel
Souza, Amauri H
contents Deep neural networks are notoriously miscalibrated, i.e., their outputs do not reflect the true probability of the event we aim to predict. While networks for tabular or image data are usually overconfident, recent works have shown that graph neural networks (GNNs) show the opposite behavior for node-level classification. But what happens when we are predicting links? We show that, in this case, GNNs often exhibit a mixed behavior. More specifically, they may be overconfident in negative predictions while being underconfident in positive ones. Based on this observation, we propose IN-N-OUT, the first-ever method to calibrate GNNs for link prediction. IN-N-OUT is based on two simple intuitions: i) attributing true/false labels to an edge while respecting a GNNs prediction should cause but small fluctuations in that edge's embedding; and, conversely, ii) if we label that same edge contradicting our GNN, embeddings should change more substantially. An extensive experimental campaign shows that IN-N-OUT significantly improves the calibration of GNNs in link prediction, consistently outperforming the baselines available -- which are not designed for this specific task.
format Preprint
id arxiv_https___arxiv_org_abs_2403_04605
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle In-n-Out: Calibrating Graph Neural Networks for Link Prediction
Nascimento, Erik
Mesquita, Diego
Kaski, Samuel
Souza, Amauri H
Machine Learning
Deep neural networks are notoriously miscalibrated, i.e., their outputs do not reflect the true probability of the event we aim to predict. While networks for tabular or image data are usually overconfident, recent works have shown that graph neural networks (GNNs) show the opposite behavior for node-level classification. But what happens when we are predicting links? We show that, in this case, GNNs often exhibit a mixed behavior. More specifically, they may be overconfident in negative predictions while being underconfident in positive ones. Based on this observation, we propose IN-N-OUT, the first-ever method to calibrate GNNs for link prediction. IN-N-OUT is based on two simple intuitions: i) attributing true/false labels to an edge while respecting a GNNs prediction should cause but small fluctuations in that edge's embedding; and, conversely, ii) if we label that same edge contradicting our GNN, embeddings should change more substantially. An extensive experimental campaign shows that IN-N-OUT significantly improves the calibration of GNNs in link prediction, consistently outperforming the baselines available -- which are not designed for this specific task.
title In-n-Out: Calibrating Graph Neural Networks for Link Prediction
topic Machine Learning
url https://arxiv.org/abs/2403.04605