Cross-property deep transfer learning framework for enhanced predictive analytics on small materials data

Journal Article · 14 November 2021 · Nature Communications

DOI: https://doi.org/10.1038/s41467-021-26921-5 · OSTI ID:1830279

Gupta, Vishu; Choudhary, Kamal ; Tavazza, Francesca ; Campbell, Carelyn ; Liao, Wei-keng; Choudhary, Alok; Agrawal, Ankit

Abstract Artificial intelligence (AI) and machine learning (ML) have been increasingly used in materials science to build predictive models and accelerate discovery. For selected properties, availability of large databases has also facilitated application of deep learning (DL) and transfer learning (TL). However, unavailability of large datasets for a majority of properties prohibits widespread application of DL/TL. We present a cross-property deep-transfer-learning framework that leverages models trained on large datasets to build models on small datasets of different properties. We test the proposed framework on 39 computational and two experimental datasets and find that the TL models with only elemental fractions as input outperform ML/DL models trained from scratch even when they are allowed to use physical attributes as input, for 27/39 (≈ 69%) computational and both the experimental datasets. We believe that the proposed framework can be widely useful to tackle the small data challenge in applying AI/ML in materials science.

• Research Organization: Northwestern University, Evanston, IL (United States)
• Sponsoring Organization: National Institute of Standards and Technology (NIST); USDOE; USDOE Office of Science (SC)
• Grant/Contract Number: SC0014330; SC0019358; SC0021399
• OSTI ID: 1830279
• Journal Information: Nature Communications, Journal Name: Nature Communications Journal Issue: 1 Vol. 12; ISSN 2041-1723
• Publisher: Nature Publishing GroupCopyright Statement
• Country of Publication: United Kingdom
• Language: English