CaltechAUTHORS
  A Caltech Library Service

Materials representation and transfer learning for multi-property prediction

Kong, Shufeng and Guevarra, Dan and Gomes, Carla P. and Gregoire, John M. (2021) Materials representation and transfer learning for multi-property prediction. Applied Physics Reviews, 8 (2). Art. No. 021409. ISSN 1931-9401. doi:10.1063/5.0047066. https://resolver.caltech.edu/CaltechAUTHORS:20210626-225301174

[img] PDF - Published Version
Creative Commons Attribution.

2MB
[img] PDF - Accepted Version
Creative Commons Attribution.

4MB
[img] PDF - Submitted Version
Creative Commons Attribution Non-commercial.

4MB

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20210626-225301174

Abstract

The adoption of machine learning in materials science has rapidly transformed materials property prediction. Hurdles limiting full capitalization of recent advancements in machine learning include the limited development of methods to learn the underlying interactions of multiple elements as well as the relationships among multiple properties to facilitate property prediction in new composition spaces. To address these issues, we introduce the Hierarchical Correlation Learning for Multi-property Prediction (H-CLMP) framework that seamlessly integrates: (i) prediction using only a material's composition, (ii) learning and exploitation of correlations among target properties in multi-target regression, and (iii) leveraging training data from tangential domains via generative transfer learning. The model is demonstrated for prediction of spectral optical absorption of complex metal oxides spanning 69 three-cation metal oxide composition spaces. H-CLMP accurately predicts non-linear composition-property relationships in composition spaces for which no training data are available, which broadens the purview of machine learning to the discovery of materials with exceptional properties. This achievement results from the principled integration of latent embedding learning, property correlation learning, generative transfer learning, and attention models. The best performance is obtained using H-CLMP with transfer learning [H-CLMP(T)] wherein a generative adversarial network is trained on computational density of states data and deployed in the target domain to augment prediction of optical absorption from composition. H-CLMP(T) aggregates multiple knowledge sources with a framework that is well suited for multi-target regression across the physical sciences.


Item Type:Article
Related URLs:
URLURL TypeDescription
https://doi.org/10.1063/5.0047066DOIArticle
https://arxiv.org/abs/2106.02225arXivDiscussion Paper
https://doi.org/10.26434/chemrxiv.14612307.v1DOIDiscussion Paper
https://data.caltech.edu/records/1878Related ItemData
https://www.cs.cornell.edu/gomes/udiscoverit/?tag=materialsRelated Itemsource code and additional data for H-CLMP
https://github.com/gomes-lab/H-CLMPRelated Itemsource code for H-CLMP and the cWGAN for transfer learning
ORCID:
AuthorORCID
Guevarra, Dan0000-0002-9592-3195
Gomes, Carla P.0000-0002-4441-7225
Gregoire, John M.0000-0002-2863-5265
Additional Information:© 2021 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). Submitted: 9 February 2021. Accepted: 28 May 2021. Published Online: 23 June 2021. This paper is part of the special collection on Autonomous (AI-driven) Materials ScienceThis paper is part of the special collection on Autonomous (AI-driven) Materials Science. This work was funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Award DE-SC0020383 (data curation, design of multi-property prediction setting and transfer setting, model evaluation) and by the Toyota Research Institute through the Accelerated Materials Design and Discovery program (development of machine learning models). The authors thank Santosh K. Suram for assistance with curation of the dataset. DATA AVAILABILITY. The data that support the findings of this study are available at https://data.caltech.edu/records/1878. The source code and additional data for H-CLMP are available at https://www.cs.cornell.edu/gomes/udiscoverit/?tag=materials. The source code for H-CLMP and the cWGAN for transfer learning are also available at https://github.com/gomes-lab/H-CLMP.
Group:JCAP
Funders:
Funding AgencyGrant Number
Department of Energy (DOE)DE-SC0020383
Toyota Research InstituteUNSPECIFIED
Issue or Number:2
DOI:10.1063/5.0047066
Record Number:CaltechAUTHORS:20210626-225301174
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20210626-225301174
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:109615
Collection:CaltechAUTHORS
Deposited By: George Porter
Deposited On:28 Jun 2021 22:19
Last Modified:01 Jun 2023 23:00

Repository Staff Only: item control page