Uncertainty for Active Learning on Graphs

Dominik Fuchsgruber, Tom Wollschläger, Bertrand Charpentier, Antonio Oroz, Stephan Günnemann

Keywords: Active Learning, Uncertainty Estimation, Transductive, Graph, Node Classification, Bayesian

TL;DR: We derive ground-truth uncertainties for node classification, prove that epistemic uncertainty sampling is an optimal active learning strategy, and show the effectiveness of our results on real data by proposing a novel estimator.

Abstract: Uncertainty Sampling is an Active Learning strategy that aims to improve the data efficiency of machine learning models by iteratively acquiring labels of data points with the highest uncertainty. While it has proven effective for independent data its applicability to graphs remains under-explored. We propose the first extensive study of Uncertainty Sampling for node classification: **(1)** We benchmark Uncertainty Sampling beyond predictive uncertainty and highlight a significant performance gap to other Active Learning strategies. **(2)** We develop ground-truth Bayesian uncertainty estimates in terms of the data generating process and prove their effectiveness in guiding Uncertainty Sampling toward optimal queries. We confirm our results on synthetic data and design an approximate approach that consistently outperforms other uncertainty estimators on real datasets. **(3)** Based on this analysis, we relate pitfalls in modeling uncertainty to existing methods. Our analysis enables and informs the development of principled uncertainty estimation on graphs.

Submission Type: Extended abstract (max 4 main pages).

Submission Number: 111