MetAL: Active Semi-Supervised Learning on Graphs via Meta Learning
This addresses the challenge of efficient labeling for graph classification, which is incremental as it builds on meta-learning principles for a specific data type.
The paper tackles the problem of active learning on graph-structured data, where existing methods perform poorly, by proposing MetAL, a meta-learning-based approach that selects unlabeled instances to directly improve future model performance, and demonstrates it outperforms state-of-the-art algorithms across multiple graph datasets.
The objective of active learning (AL) is to train classification models with less number of labeled instances by selecting only the most informative instances for labeling. The AL algorithms designed for other data types such as images and text do not perform well on graph-structured data. Although a few heuristics-based AL algorithms have been proposed for graphs, a principled approach is lacking. In this paper, we propose MetAL, an AL approach that selects unlabeled instances that directly improve the future performance of a classification model. For a semi-supervised learning problem, we formulate the AL task as a bilevel optimization problem. Based on recent work in meta-learning, we use the meta-gradients to approximate the impact of retraining the model with any unlabeled instance on the model performance. Using multiple graph datasets belonging to different domains, we demonstrate that MetAL efficiently outperforms existing state-of-the-art AL algorithms.