A Data Ecosystem to Support Machine Learning in Materials Science

Abstract

Facilitating the application of machine learning to materials science problems will require enhancing the data ecosystem to enable discovery and collection of data from many sources, automated dissemination of new data across the ecosystem, and the connecting of data with materials-specific machine learning models. Here, we present two projects, the Materials Data Facility (MDF) and the Data and Learning Hub for Science (DLHub), that address these needs. We use examples to show how MDF and DLHub capabilities can be leveraged to link data with machine learning models and how users can access those capabilities through web and programmatic interfaces.

arXiv comment23 pages, 6 figures, submitted to MRS Communications special issue on AI in Materials Science

References in corpus (4)

• SchNet - a deep learning architecture for molecules and materials
• Big Data meets Quantum Chemistry Approximations: The $\Delta$-Machine Learning Approach
• Real-time coherent diffraction inversion using deep generative networks
• Machine Learning Prediction of Accurate Atomization Energies of Organic Molecules from Low-Fidelity Quantum Chemical Calculations

Cited by in corpus (44)

• A practical guide to machine learning interatomic potentials -- Status and future
• Discovery of Self-Assembling $\pi$-Conjugated Peptides by Active Learning-Directed Coarse-Grained Molecular Simulation
• FAIR for AI: An interdisciplinary and international community building perspective
• A Deep Learning Based Automatic Defect Analysis Framework for In-situ TEM Ion Irradiations
• Convergence of Artificial Intelligence and High Performance Computing on NSF-supported Cyberinfrastructure
• Tailoring spin wave channels in a reconfigurable artificial spin ice
• Colmena: Scalable Machine-Learning-Based Steering of Ensemble Simulations for High Performance Computing
• FAIR principles for AI models with a practical application for accelerated high energy diffraction microscopy
• Applications of physics informed neural operators
• An Automated Scanning Transmission Electron Microscope Guided by Sparse Data Analytics
• A Combined First Principles Study of the Structural, Magnetic, and Phonon Properties of Monolayer CrI${}_3$
• Cohesion and excitations of diamond-structure silicon by quantum Monte Carlo: Benchmarks and control of systematic biases
• Temporally coherent backmapping of molecular trajectories from coarse-grained to atomistic resolution
• Towards DMC accuracy across chemical space with scalable $\Delta$-QML
• Physics-inspired deep learning to characterize the signal manifold of quasi-circular, spinning, non-precessing binary black hole mergers
• A new generation of effective core potentials from correlated and spin-orbit calculations: selected heavy elements
• Towards Quantum Monte Carlo Forces on Heavier Ions: Scaling Properties
• Orbital character of the spin-reorientation transition in TbMn${}_6$Sn${}_6$
• Targeting SARS-CoV-2 with AI- and HPC-enabled Lead Generation: A First Data Release
• Many-body electronic structure of LaScO${}_3$ by real space quantum Monte Carlo
• Range-Separated Hybrid Functionals for Mixed-Dimensional Heterojunctions: Application to Phthalocyanines/MoS2
• Advances in Machine and Deep Learning for Modeling and Real-time Detection of Multi-Messenger Sources
• Chiral and flat-band magnetic cluster excitations in a ferromagnetic kagome metal
• Sim2Ls: FAIR simulation workflows and data
• The role of electron correlations in the electronic structure of putative Chern magnet TbMn${}_6$Sn${}_6$
• Binding and excitations in Si${}_x$H${}_y$ molecular systems using quantum Monte Carlo
• Towards Foundation Models for Materials Science: The Open MatSci ML Toolkit
• Benchmark tests of atom segmentation deep learning models with a consistent dataset
• EXSCLAIM! -- An automated pipeline for the construction of labeled materials imaging datasets from literature
• Electronic structure of $\mathbf{\alpha }$-RuCl${}_3$ by fixed-node and fixed-phase diffusion Monte Carlo methods
• mkite: A distributed computing platform for high-throughput materials simulations
• In Situ Imaging of Parachute Textile Micro-Mechanics Under Tensile Load
• Data-driven reconstruction of spectral conductivity and chemical potential from thermoelectric transport data
• Enhanced Twist-Averaging Technique for Magnetic Metals: Applications using Quantum Monte Carlo
• Identifying materials with charge-spin physics using charge-spin susceptibility computed from first principles
• External magnetic field suppression of carbon diffusion in iron
• A new generation of effective core potentials: Selected lanthanides and heavy elements II
• Orbital Mixer: Using Atomic Orbital Features for Basis Dependent Prediction of Molecular Wavefunctions
• A quantum Monte Carlo study of systems with effective core potentials and node nonlinearities
• Statistically-informed deep learning for gravitational wave parameter estimation
• Laser-induced stress wave propagation through smooth and rough substrates
• Computing ternary liquid phase diagrams: Fe-Cu-Ni
• Assessing the accuracy of compound formation energies with quantum Monte Carlo
• Investigating the effects of local environment on nitrogen vacancies in high entropy metal nitrides