ACS Central Science Virtual Issue on Machine Learning

Ferguson, Andrew L.

doi:10.1021/acscentsci.8b00528

Cited by 15 publications

(12 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Following a number of successful applications of machine-learning methods to predict materials properties, − a recent landmark paper by Brockherde et al showed that it is also possible to predict the ground-state electron density in a way that mimics the Hohenberg–Kohn mapping between the nuclear potential and the density . A smoothed representation of the nuclear potential was used as a fingerprint to describe molecular configurations and to carry out individual predictions of the expansion coefficients of ρ( r ) represented in a plane-wave basis.…”

Section: Introductionmentioning

confidence: 99%

Transferable Machine-Learning Model of the Electron Density

et al. 2018

View full text Add to dashboard Cite

The electronic charge density plays a central role in determining the behavior of matter at the atomic scale, but its computational evaluation requires demanding electronic-structure calculations. We introduce an atom-centered, symmetry-adapted framework to machine-learn the valence charge density based on a small number of reference calculations. The model is highly transferable, meaning it can be trained on electronic-structure data of small molecules and used to predict the charge density of larger compounds with low, linear-scaling cost. Applications are shown for various hydrocarbon molecules of increasing complexity and flexibility, and demonstrate the accuracy of the model when predicting the density on octane and octatetraene after training exclusively on butane and butadiene. This transferable, data-driven model can be used to interpret experiments, accelerate electronic structure calculations, and compute electrostatic interactions in molecules and condensed-phase systems.

show abstract

Section: Introductionmentioning

confidence: 99%

Transferable Machine-Learning Model of the Electron Density

et al. 2018

View full text Add to dashboard Cite

show abstract

“…Despite major issues with GAN, which are mode collapse, non-convergence, and training instability, 22 GAN has been one of the most interesting ideas in machine learning (ML) of the past 10 years. 13 Although conventional ML approaches based on supervised learning are well established in the materials research community, [23][24][25][26][27][28][29][30][31][32] GAN algorithms have just begun to be used for the materials research.…”

mentioning

confidence: 99%

“…One neural network, called the generator, generates new data instances from randomly distributed sample data, and the other, the discriminator, judges them for authenticity; in other words, the discriminator determines whether each instance of data it examines belongs to the genuine training dataset or not. In contrast to the recent boom for deep learning-based artificial intelligence for use in both the chemistry and materials science research fields, [23][24][25][26][27][28][29][30][31][32] the unsupervised learning-based GAN is yet to be spotlighted in both the fields. Nevertheless, we have seen a certain degree of progress in the GAN utilization for the design of molecules, [34][35][36][37][38][39] and the drug discovery, [40][41][42] the inverse design of materials, 43 the design of crystal structure of inorganic materials.…”

mentioning

confidence: 99%

Virtual microstructure design for steels using generative adversarial networks

Lee

Goo

Park

et al. 2020

Engineering Reports

View full text Add to dashboard Cite

The prediction of macro‐scale materials properties from microstructures, and vice versa, should be a key part in modeling quantitative microstructure‐physical property relationships. It would be helpful if the microstructural input and output were in the form of visual images rather than parameterized descriptors. However, only a typical supervised learning technique would be insufficient to build up a model with real‐image‐output. A generative adversarial network (GAN) is required to treat visual images as output for a promising PMPR model. Recently developed deep‐learning‐based GAN techniques such as a deep convolutional GAN (DCGAN), a cycle‐consistent GAN (Cycle GAN), and a conditional GAN‐based image to image translation (Pix2Pix) could be of great help via the creation of realistic microstructures. In this regard, we generated virtual micrographs for various types of steels using a DCGAN, a Cycle GAN, and a Pix2Pix and confirmed the generated micrographs are qualitatively indistinguishable from the ground truth.

show abstract

“…The integration with experimental results is yet another longterm challenge. Finally, the combination of databases with machine learning is having a big impact on the field by allowing an increasing number of applications 33,[38][39][40][41][42] and is seen as a major opportunity in industry 43 . However, this blooming computational field also demands massive curated data and robust algorithm benchmarks 44 , as there is a risk of the hype masking the real advantages of these powerful tools 45,46 .…”

mentioning

confidence: 99%