Semi-Supervised Learning for Molecular Graphs via Ensemble Consensus

2026-07-30

AI Summary

This work addresses the challenges of scarce labeled data, high annotation costs, and the difficulty of designing label-preserving data augmentation in molecular graphs by proposing an ensemble-based semi-supervised learning paradigm grounded in consistency regularization. The method fosters knowledge collaboration among multiple graph neural networks, achieving a knowledge distillation–like effect without requiring explicit data augmentation. Evaluated across diverse molecular datasets, task types, and model architectures, the proposed approach consistently yields significant performance gains: a single model trained with this framework surpasses conventional fully supervised ensembles in accuracy, while simultaneously enhancing robustness and reducing calibration error.

Abstract

Machine learning is transforming molecular sciences by accelerating property prediction, simulation, and the discovery of new molecules and materials. Acquiring labeled data in these domains is often costly and time-consuming, whereas large collections of unlabeled molecular data are readily available. Standard semi-supervised learning methods often rely on label-preserving augmentations, which are challenging to design in the molecular domain, where minor changes can drastically alter properties. In this work, we show that semi-supervised methods that rely on an ensemble consensus can boost predictive accuracy across a diverse range of molecular datasets, task types, and graph neural network architectures. We find that training with an ensemble consensus objective increases robustness in models and exhibits an effect similar to knowledge distillation; an individual member of an ensemble trained this way outperforms a full ensemble trained in a traditional supervised fashion in almost all cases. In addition, this type of semi-supervised training reduces calibration error.

Problem

• Semi-Supervised Learning
• Molecular Graphs
• Label Scarcity
• Data Augmentation
• Property Prediction

Innovation

• semi-supervised learning
• ensemble consensus
• molecular graphs
• graph neural networks
• knowledge distillation