Machine learning identifies scale-free properties in disordered materials

Abstract

The vast amount of design freedom in disordered systems expands the parameter space for signal processing, allowing for unique signal flows that are distinguished from those in regular systems. However, this large degree of freedom has hindered the deterministic design of disordered systems for target functionalities. Here, we employ a machine learning (ML) approach for predicting and designing wave-matter interactions in disordered structures, thereby identifying scale-free properties for waves. To abstract and map the features of wave behaviours and disordered structures, we develop disorder-to-localization and localization-to-disorder convolutional neural networks (CNNs). Each CNN enables the instantaneous prediction of wave localization in disordered structures and the instantaneous generation of disordered structures from given localizations. We demonstrate that CNN-generated disordered structures have scale-free properties with heavy tails and hub atoms, which exhibit an increase of multiple orders of magnitude in robustness to accidental defects, such as material or structural imperfection. Our results verify the critical role of ML network structures in determining ML-generated real-space structures, which can be used in the design of defect-immune and efficiently tunable devices.

arXiv comment44 pages, 15 figures

References in corpus (28)

• Emergence of scaling in random networks
• TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
• Power-law distributions in empirical data
• Machine learning and the physical sciences
• Machine learning phases of matter
• Breakdown of the Internet under intentional attack
• Training deep neural networks for the inverse design of nanophotonic structures
• Phase recovery and holographic image reconstruction using deep learning in neural networks
• Powerlaw: a Python package for analysis of heavy-tailed distributions
• A Generative Model for Inverse Design of Metamaterials
• Local Density Fluctuations, Hyperuniformity, and Order Metrics
• Hyperuniform States of Matter
• Experimental observation of the Anderson transition with atomic matter waves
• Localization and Mobility Edge in One-Dimensional Potentials with Correlated Disorder
• Machine learning quantum phases of matter beyond the fermion sign problem
• Photonic Topological Anderson Insulators
• Insightful classification of crystal structures using deep learning
• Hyperbolic Lattices in Circuit Quantum Electrodynamics
• Identifying topological order through unsupervised machine learning
• Isotropic Band Gaps and Freeform Waveguides Observed in Hyperuniform Disordered Photonic Solids
• Deep Learning the Quantum Phase Transitions in Random Two-Dimensional Electron Systems
• Topological hyperbolic lattices
• Plasmonic colours predicted by deep learning
• Broadband Coherent Enhancement of Transmission and Absorption in Disordered Media
• Drawing Phase Diagrams of Random Quantum Systems by Deep Learning the Wave Functions
• Weak localization of light in superdiffusive random systems
• Interdimensional optical isospectrality inspired by graph networks
• The Role of Power-Law Correlated Disorder in the Anderson Metal-Insulator Transition

Cited by in corpus (6)

• Heavy tails and pruning in programmable photonic circuits
• Deep learning of deformation-dependent conductance in thin films: nanobubbles in graphene
• Analysis and fabrication of a photonic crystal based anti-reflective coating for photovoltaics generated by evolutionary optimization
• Inverse-designed dispersive time-varying nanostructures
• Topological protection in nonlinear optical dynamics with parity-time symmetry
• Deep-subwavelength engineering of stealthy hyperuniformity