Discovering the Building Blocks of Atomic Systems using Machine Learning

Abstract

Machine learning has proven to be a valuable tool to approximate functions in high-dimensional spaces. Unfortunately, analysis of these models to extract the relevant physics is never as easy as applying machine learning to a large dataset in the first place. Here we present a description of atomic systems that generates machine learning representations with a direct path to physical interpretation. As an example, we demonstrate its usefulness as a universal descriptor of grain boundary systems. Grain boundaries in crystalline materials are a quintessential example of a complex, high-dimensional system with broad impact on many physical properties including strength, ductility, corrosion resistance, crack resistance, and conductivity. In addition to modeling such properties, the method also provides insight into the physical "building blocks" that influence them. This opens the way to discover the underlying physics behind behaviors by understanding which building blocks map to particular properties. Once the structures are understood, they can then be optimized for desirable behaviors.

8 pages, 4 figures, 1 table

References in corpus (2)

• Imbalanced-learn: A Python Toolbox to Tackle the Curse of Imbalanced Datasets in Machine Learning
• Prediction of grain boundary structure and energy by machine learning

Cited by in corpus (20)

• Machine Learning in Electronic Quantum Matter Imaging Experiments
• Advances of Machine Learning in Materials Science: Ideas and Techniques
• Examination of computed aluminum grain boundary structures and interface energies that span the 5D space of crystallographic character
• A Novel Approach to Describe Chemical Environments in High Dimensional Neural Network Potentials
• Can Symmetric Tilt Grain Boundaries Represent Polycrystals?
• A Physically-informed Graph-based Order Parameter for the Universal Characterization of Atomic Structures
• TimeSOAP: Tracking high-dimensional fluctuations in complex molecular systems via time-variations of SOAP spectra
• Describe, Transform, Machine Learning: Feature Engineering for Grain Boundaries and Other Variable-Sized Atom Clusters
• Quantifying Disorder One Atom at a Time Using an Interpretable Graph Neural Network Paradigm
• Knowledge-Transfer based Cost-effective Search for Interface Structures: A Case Study on fcc-Al [110] Tilt Grain Boundary
• Can we predict mixed grain boundaries from their tilt and twist components?
• Structures and energies of computed silicon (001) small angle mixed grain boundaries as a function of three macroscopic characters
• Maximum volume simplex method for automatic selection and classification of atomic environments and environment descriptor compression
• Facet and energy predictions in grain boundaries: lattice matching and molecular dynamics
• Classification of magnetic order from electronic structure by using machine learning
• Structural Characterization of Grain Boundaries and Machine Learning of Grain Boundary Energy and Mobility
• Learning from metastable grain boundaries
• RAFFLE: Active learning accelerated interface structure prediction
• Robust recognition and exploratory analysis of crystal structures via Bayesian deep learning
• Gap prediction in hybrid graphene - hexagonal boron nitride nanoflakes using artificial neural networks