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

Title: A Statistical Learning Framework for Materials Science: Application to Elastic Moduli of k-nary Inorganic Polycrystalline Compounds

Abstract

Materials scientists increasingly employ machine or statistical learning (SL) techniques to accelerate materials discovery and design. Such pursuits benefit from pooling training data across, and thus being able to generalize predictions over, k-nary compounds of diverse chemistries and structures. This work presents a SL framework that addresses challenges in materials science applications, where datasets are diverse but of modest size, and extreme values are often of interest. Our advances include the application of power or Hölder means to construct descriptors that generalize over chemistry and crystal structure, and the incorporation of multivariate local regression within a gradient boosting framework. The approach is demonstrated by developing SL models to predict bulk and shear moduli (K and G, respectively) for polycrystalline inorganic compounds, using 1,940 compounds from a growing database of calculated elastic moduli for metals, semiconductors and insulators. The usefulness of the models is illustrated by screening for superhard materials.

Authors:
 [1];  [2];  [3];  [2];  [4];  [2];  [4];  [3]
  1. Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States). Dept. of Materials Science and Engineering
  2. Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States). Energy Technologies Area
  3. Univ. of California, San Diego, CA (United States). Computational and Applied Statistics Laboratory
  4. Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States). Materials Sciences Division
Publication Date:
Research Org.:
Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States)
Sponsoring Org.:
USDOE Office of Science (SC), Basic Energy Sciences (BES) (SC-22)
OSTI Identifier:
1377526
Grant/Contract Number:  
AC02-05CH11231; EDCBEE
Resource Type:
Journal Article: Accepted Manuscript
Journal Name:
Scientific Reports
Additional Journal Information:
Journal Volume: 6; Journal Issue: 1; Journal ID: ISSN 2045-2322
Publisher:
Nature Publishing Group
Country of Publication:
United States
Language:
English
Subject:
36 MATERIALS SCIENCE

Citation Formats

de Jong, Maarten, Chen, Wei, Notestine, Randy, Persson, Kristin, Ceder, Gerbrand, Jain, Anubhav, Asta, Mark, and Gamst, Anthony. A Statistical Learning Framework for Materials Science: Application to Elastic Moduli of k-nary Inorganic Polycrystalline Compounds. United States: N. p., 2016. Web. doi:10.1038/srep34256.
de Jong, Maarten, Chen, Wei, Notestine, Randy, Persson, Kristin, Ceder, Gerbrand, Jain, Anubhav, Asta, Mark, & Gamst, Anthony. A Statistical Learning Framework for Materials Science: Application to Elastic Moduli of k-nary Inorganic Polycrystalline Compounds. United States. doi:10.1038/srep34256.
de Jong, Maarten, Chen, Wei, Notestine, Randy, Persson, Kristin, Ceder, Gerbrand, Jain, Anubhav, Asta, Mark, and Gamst, Anthony. Mon . "A Statistical Learning Framework for Materials Science: Application to Elastic Moduli of k-nary Inorganic Polycrystalline Compounds". United States. doi:10.1038/srep34256. https://www.osti.gov/servlets/purl/1377526.
@article{osti_1377526,
title = {A Statistical Learning Framework for Materials Science: Application to Elastic Moduli of k-nary Inorganic Polycrystalline Compounds},
author = {de Jong, Maarten and Chen, Wei and Notestine, Randy and Persson, Kristin and Ceder, Gerbrand and Jain, Anubhav and Asta, Mark and Gamst, Anthony},
abstractNote = {Materials scientists increasingly employ machine or statistical learning (SL) techniques to accelerate materials discovery and design. Such pursuits benefit from pooling training data across, and thus being able to generalize predictions over, k-nary compounds of diverse chemistries and structures. This work presents a SL framework that addresses challenges in materials science applications, where datasets are diverse but of modest size, and extreme values are often of interest. Our advances include the application of power or Hölder means to construct descriptors that generalize over chemistry and crystal structure, and the incorporation of multivariate local regression within a gradient boosting framework. The approach is demonstrated by developing SL models to predict bulk and shear moduli (K and G, respectively) for polycrystalline inorganic compounds, using 1,940 compounds from a growing database of calculated elastic moduli for metals, semiconductors and insulators. The usefulness of the models is illustrated by screening for superhard materials.},
doi = {10.1038/srep34256},
journal = {Scientific Reports},
number = 1,
volume = 6,
place = {United States},
year = {Mon Oct 03 00:00:00 EDT 2016},
month = {Mon Oct 03 00:00:00 EDT 2016}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Citation Metrics:
Cited by: 22 works
Citation information provided by
Web of Science

Save / Share:

Works referenced in this record:

Complex thermoelectric materials
journal, February 2008

  • Snyder, G. Jeffrey; Toberer, Eric S.
  • Nature Materials, Vol. 7, Issue 2, p. 105-114
  • DOI: 10.1038/nmat2090

Lower limit to the thermal conductivity of disordered crystals
journal, September 1992

  • Cahill, David G.; Watson, S. K.; Pohl, R. O.
  • Physical Review B, Vol. 46, Issue 10, p. 6131-6140
  • DOI: 10.1103/PhysRevB.46.6131