Machine learning-assisted design of material properties

Abstract

Designing functional materials requires a deep search through multidimensional spaces for system parameters that yield desirable material properties. For cases where conventional parameter sweeps or trial-and-error sampling are impractical, inverse methods that frame design as a constrained optimization problem present an attractive alternative. However, even efficient algorithms require time- and resource-intensive characterization of material properties many times during optimization, imposing a design bottleneck. Approaches that incorporate machine learning can help address this limitation and accelerate the discovery of materials with targeted properties. In this article, we review how to leverage machine learning to reduce dimensionality in order to effectively explore design space, accelerate property evaluation, and generate unconventional material structures with optimal properties. We also discuss promising future directions, including integration of machine learning into multiple stages of a design algorithm and interpretation of machine learning models to understand how design parameters relate to material properties.

arXiv comment43 pages, 4 figures, accepted for publication in Annual Review of Chemical and Biomolecular Engineering

References in corpus (27)

• Auto-Encoding Variational Bayes
• Generative Adversarial Networks
• Automatic chemical design using a data-driven continuous representation of molecules
• Axiomatic Attribution for Deep Networks
• Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
• Molecular Graph Convolutions: Moving Beyond Fingerprints
• Training deep neural networks for the inverse design of nanophotonic structures
• A Generative Model for Inverse Design of Metamaterials
• SISSO: a compressed-sensing method for identifying the best low-dimensional descriptor in an immensity of offered candidates
• Deep learning for molecular design - a review of the state of the art
• AtomNet: A Deep Convolutional Neural Network for Bioactivity Prediction in Structure-based Drug Discovery
• Global optimization of dielectric metasurfaces using a physics-driven neural network
• Freeform Diffractive Metagrating Design Based on Generative Adversarial Networks
• Deep Generative Modeling for Mechanistic-based Learning and Design of Metamaterial Systems
• Machine-Learning-Assisted Metasurface Design for High-Efficiency Thermal Emitter Optimization
• Deep learning approach based on dimensionality reduction for designing electromagnetic nanostructures
• Autonomous discovery of battery electrolytes with robotic experimentation and machine-learning
• Graph Neural Networks for an Accurate and Interpretable Prediction of the Properties of Polycrystalline Materials
• Mapping the global design space of nanophotonic components using machine learning pattern recognition
• Scaffold-based molecular design using graph generative model
• Generative Models for Automatic Chemical Design
• Inverse methods for design of soft materials
• Perspective: Inverse methods for material design
• Deeply learning molecular structure-property relationships using attention- and gate-augmented graph convolutional network
• Interpretable deep learning for guided structure-property explorations in photovoltaics
• Unsupervised machine learning for detection of phase transitions in off-lattice systems I. Foundations
• Unsupervised machine learning for detection of phase transitions in off-lattice systems II. Applications

Cited by in corpus (5)

• Plasmonic Metal Oxide Nanocrystals as Building Blocks for Infrared Metasurfaces
• Geometric model of crack-templated networks for transparent conductive films
• Plasmon Polaritons in Disordered Nanoparticle Assemblies
• Emerging Microelectronic Materials by Design: Navigating Combinatorial Design Space with Scarce and Dispersed Data
• relentless: Transparent, reproducible molecular dynamics simulations for optimization