skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Identifying Structural Flow Defects in Disordered Solids Using Machine-Learning Methods

; ; ; ; ; ; ;
Publication Date:
Sponsoring Org.:
OSTI Identifier:
Grant/Contract Number:
Resource Type:
Journal Article: Publisher's Accepted Manuscript
Journal Name:
Physical Review Letters
Additional Journal Information:
Journal Volume: 114; Journal Issue: 10; Journal ID: ISSN 0031-9007
American Physical Society
Country of Publication:
United States

Citation Formats

Cubuk, E. D., Schoenholz, S. S., Rieser, J. M., Malone, B. D., Rottler, J., Durian, D. J., Kaxiras, E., and Liu, A. J. Identifying Structural Flow Defects in Disordered Solids Using Machine-Learning Methods. United States: N. p., 2015. Web. doi:10.1103/PhysRevLett.114.108001.
Cubuk, E. D., Schoenholz, S. S., Rieser, J. M., Malone, B. D., Rottler, J., Durian, D. J., Kaxiras, E., & Liu, A. J. Identifying Structural Flow Defects in Disordered Solids Using Machine-Learning Methods. United States. doi:10.1103/PhysRevLett.114.108001.
Cubuk, E. D., Schoenholz, S. S., Rieser, J. M., Malone, B. D., Rottler, J., Durian, D. J., Kaxiras, E., and Liu, A. J. 2015. "Identifying Structural Flow Defects in Disordered Solids Using Machine-Learning Methods". United States. doi:10.1103/PhysRevLett.114.108001.
title = {Identifying Structural Flow Defects in Disordered Solids Using Machine-Learning Methods},
author = {Cubuk, E. D. and Schoenholz, S. S. and Rieser, J. M. and Malone, B. D. and Rottler, J. and Durian, D. J. and Kaxiras, E. and Liu, A. J.},
abstractNote = {},
doi = {10.1103/PhysRevLett.114.108001},
journal = {Physical Review Letters},
number = 10,
volume = 114,
place = {United States},
year = 2015,
month = 3

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record at 10.1103/PhysRevLett.114.108001

Citation Metrics:
Cited by: 45works
Citation information provided by
Web of Science

Save / Share:
  • Of all the activity observed on the Sun, two of the most energetic events are flares and coronal mass ejections (CMEs). Usually, solar active regions that produce large flares will also produce a CME, but this is not always true. Despite advances in numerical modeling, it is still unclear which circumstances will produce a CME. Therefore, it is worthwhile to empirically determine which features distinguish flares associated with CMEs from flares that are not. At this time, no extensive study has used physically meaningful features of active regions to distinguish between these two populations. As such, we attempt to domore » so by using features derived from (1) photospheric vector magnetic field data taken by the Solar Dynamics Observatory ’s Helioseismic and Magnetic Imager instrument and (2) X-ray flux data from the Geostationary Operational Environmental Satellite’s X-ray Flux instrument. We build a catalog of active regions that either produced both a flare and a CME (the positive class) or simply a flare (the negative class). We then use machine-learning algorithms to (1) determine which features distinguish these two populations, and (2) forecast whether an active region that produces an M- or X-class flare will also produce a CME. We compute the True Skill Statistic, a forecast verification metric, and find that it is a relatively high value of ∼0.8 ± 0.2. We conclude that a combination of six parameters, which are all intensive in nature, will capture most of the relevant information contained in the photospheric magnetic field.« less
  • A machine learning–based framework for modeling the error introduced by surrogate models of parameterized dynamical systems is proposed. The framework entails the use of high-dimensional regression techniques (eg, random forests, and LASSO) to map a large set of inexpensively computed “error indicators” (ie, features) produced by the surrogate model at a given time instance to a prediction of the surrogate-model error in a quantity of interest (QoI). This eliminates the need for the user to hand-select a small number of informative features. The methodology requires a training set of parameter instances at which the time-dependent surrogate-model error is computed bymore » simulating both the high-fidelity and surrogate models. Using these training data, the method first determines regression-model locality (via classification or clustering) and subsequently constructs a “local” regression model to predict the time-instantaneous error within each identified region of feature space. We consider 2 uses for the resulting error model: (1) as a correction to the surrogate-model QoI prediction at each time instance and (2) as a way to statistically model arbitrary functions of the time-dependent surrogate-model error (eg, time-integrated errors). We then apply the proposed framework to model errors in reduced-order models of nonlinear oil-water subsurface flow simulations, with time-varying well-control (bottom-hole pressure) parameters. The reduced-order models used in this work entail application of trajectory piecewise linearization in conjunction with proper orthogonal decomposition. Moreover, when the first use of the method is considered, numerical experiments demonstrate consistent improvement in accuracy in the time-instantaneous QoI prediction relative to the original surrogate model, across a large number of test cases. When the second use is considered, results show that the proposed method provides accurate statistical predictions of the time- and well-averaged errors.« less
  • Purpose: Given the paucity of available data concerning radiotherapy-induced urinary toxicity, it is important to ensure derivation of the most robust models with superior predictive performance. This work explores multiple statistical-learning strategies for prediction of urinary symptoms following external beam radiotherapy of the prostate. Methods: The performance of logistic regression, elastic-net, support-vector machine, random forest, neural network, and multivariate adaptive regression splines (MARS) to predict urinary symptoms was analyzed using data from 754 participants accrued by TROG03.04-RADAR. Predictive features included dose-surface data, comorbidities, and medication-intake. Four symptoms were analyzed: dysuria, haematuria, incontinence, and frequency, each with three definitions (grade ≥more » 1, grade ≥ 2 and longitudinal) with event rate between 2.3% and 76.1%. Repeated cross-validations producing matched models were implemented. A synthetic minority oversampling technique was utilized in endpoints with rare events. Parameter optimization was performed on the training data. Area under the receiver operating characteristic curve (AUROC) was used to compare performance using sample size to detect differences of ≥0.05 at the 95% confidence level. Results: Logistic regression, elastic-net, random forest, MARS, and support-vector machine were the highest-performing statistical-learning strategies in 3, 3, 3, 2, and 1 endpoints, respectively. Logistic regression, MARS, elastic-net, random forest, neural network, and support-vector machine were the best, or were not significantly worse than the best, in 7, 7, 5, 5, 3, and 1 endpoints. The best-performing statistical model was for dysuria grade ≥ 1 with AUROC ± standard deviation of 0.649 ± 0.074 using MARS. For longitudinal frequency and dysuria grade ≥ 1, all strategies produced AUROC>0.6 while all haematuria endpoints and longitudinal incontinence models produced AUROC<0.6. Conclusions: Logistic regression and MARS were most likely to be the best-performing strategy for the prediction of urinary symptoms with elastic-net and random forest producing competitive results. The predictive power of the models was modest and endpoint-dependent. New features, including spatial dose maps, may be necessary to achieve better models.« less
  • A new approach to the investigation of magneto-structural correlations in solids containing antiferromagnetically coupled transition-metal centers is described that illustrates the potential of NMR spectroscopy in such work. The results of a variable-temperature (VT) /sup 13/C cross-polarization magic-angle-spinning (CP/MAS) NMR investigation of anhydrous copper(II) n-butyrate, (Cu(C/sub 3/H/sub 7/COO)/sub 2/)/sub 2/ are reported. Isotropic shifts are found to be primarily contact in origin, and a statistical analysis of their temperature dependence allows the calculation of singlet-triplet energy level separations (-2J), diamagnetic shifts (delta/sub dia/), and electron-nucleus hyperfine coupling constants (A), which are shown to give insight into the mechanisms of electronmore » delocalization along the superexchange pathway. Signal multiplicity can be related to compound structure, which was determined by using x-ray crystallography. The title compound is triclinic and has a space group of P/anti 1/ with a = 9.035 (2) /angstrom/, b = 5.192 (2) /angstrom/, c = 11.695 (3) /angstrom/, ..cap alpha.. = 85.88 (2)/degrees/, ..gamma.. = 109.32 (2)/degrees/, Z = 1, and V = 515.2 (3) /angstrom//sup 3/; the final weighted R value for 2169 reflections was 0.048. 21 references, 4 figures, 4 tables.« less
  • Cited by 11