Title: Reliable and explainable machine-learning methods for accelerated material discovery

Abstract

Abstract Despite ML’s impressive performance in commercial applications, several unique challenges exist when applying ML in materials science applications. In such a context, the contributions of this work are twofold. First, we identify common pitfalls of existing ML techniques when learning from underrepresented/imbalanced material data. Specifically, we show that with imbalanced data, standard methods for assessing quality of ML models break down and lead to misleading conclusions. Furthermore, we find that the model’s own confidence score cannot be trusted and model introspection methods (using simpler models) do not help as they result in loss of predictive performance (reliability-explainability trade-off). Second, to overcome these challenges, we propose a general-purpose explainable and reliable machine-learning framework. Specifically, we propose a generic pipeline that employs an ensemble of simpler models to reliably predict material properties. We also propose a transfer learning technique and show that the performance loss due to models’ simplicity can be overcome by exploiting correlations among different material properties. A new evaluation metric and a trust score to better quantify the confidence in the predictions are also proposed. To improve the interpretability, we add a rationale generator component to our framework which provides both model-level and decision-level explanations. Finally, we demonstratemore » the versatility of our technique on two applications: (1) predicting properties of crystalline compounds and (2) identifying potentially stable solar cell materials. We also point to some outstanding issues yet to be resolved for a successful application of ML in material science.« less

Authors:

Kailkhura, Bhavya; Gallagher, Brian; Kim, Sookyung; Hiszpanski, Anna; Han, T. Yong-Jin

• Search DOE PAGES for author "Han, T. Yong-Jin"
• Search DOE PAGES for ORCID "0000-0002-3000-2782"

Publication Date:

Thu Nov 14 00:00:00 EST 2019

Research Org.:

Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)

Sponsoring Org.:

USDOE National Nuclear Security Administration (NNSA); USDOE Laboratory Directed Research and Development (LDRD) Program

OSTI Identifier:

1619662

Alternate Identifier(s):

OSTI ID: 1734611

Report Number(s):

LLNL-JRNL-764864
Journal ID: ISSN 2057-3960; 108; PII: 248

Grant/Contract Number:  

AC52-07NA27344; 16-ERD-019; 19-SI-00

Resource Type:

Published Article

Journal Name:

npj Computational Materials

Additional Journal Information:

Journal Name: npj Computational Materials Journal Volume: 5 Journal Issue: 1; Journal ID: ISSN 2057-3960

Publisher:

Nature Publishing Group

Country of Publication:

United Kingdom

Language:

English

Subject:

97 MATHEMATICS AND COMPUTING; Computational methods; design, synthesis and processing