Unsupervised Atomic Data Mining via Multi-Kernel Graph Autoencoders for Machine Learning Force Fields

2025-09-15

AI Summary

In machine learning force field (MLFF) training, chemical diversity in datasets is often compromised by sampling bias across the potential energy surface (PES), while conventional clustering and pruning methods fail in high-dimensional atomic descriptor spaces. To address this, we propose MEAGraph—a novel unsupervised graph autoencoder integrating multi-kernel linear transformations and edge-wise attention mechanisms—capable of learning geometry-sensitive atomic environment representations without labeled data. MEAGraph effectively disentangles distinct PES regions, enabling accurate clustering and efficient, lightweight dataset pruning. Validation on Nb, Ta, and Fe systems demonstrates substantial improvements in chemical diversity and distributional balance of training data, leading to enhanced MLFF generalization performance. The framework establishes a scalable, label-free paradigm for constructing high-quality, chemically diverse datasets for robust force field training.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Search and Optimization: Metareasoning and Metaheuristics

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Constructing a chemically diverse dataset while avoiding sampling bias is critical to training efficient and generalizable force fields. However, in computational chemistry and materials science, many common dataset generation techniques are prone to oversampling regions of the potential energy surface. Furthermore, these regions can be difficult to identify and isolate from each other or may not align well with human intuition, making it challenging to systematically remove bias in the dataset. While traditional clustering and pruning (down-sampling) approaches can be useful for this, they can often lead to information loss or a failure to properly identify distinct regions of the potential energy surface due to difficulties associated with the high dimensionality of atomic descriptors. In this work, we introduce the Multi-kernel Edge Attention-based Graph Autoencoder (MEAGraph) model, an unsupervised approach for analyzing atomic datasets. MEAGraph combines multiple linear kernel transformations with attention-based message passing to capture geometric sensitivity and enable effective dataset pruning without relying on labels or extensive training. Demonstrated applications on niobium, tantalum, and iron datasets show that MEAGraph efficiently groups similar atomic environments, allowing for the use of basic pruning techniques for removing sampling bias. This approach provides an effective method for representation learning and clustering that can be used for data analysis, outlier detection, and dataset optimization.

Problem

• Unsupervised method identifies sampling bias in atomic datasets
• Multi-kernel graph autoencoder clusters high-dimensional atomic environments
• Prunes datasets to remove bias without energy labels

Innovation

• Unsupervised graph autoencoder for atomic data mining
• Multi-kernel attention for geometric sensitivity capture
• Label-free clustering and pruning of atomic environments