Engineering and Technology

Crystal Twins: Self-Supervised Learning for Crystalline Material Property Prediction

R. Magar, Y. Wang, et al.

Discover the groundbreaking Crystal Twins (CT) method developed by Rishikesh Magar, Yuyang Wang, and Amir Barati Farimani, which harnesses self-supervised learning to effectively predict material properties using large unlabeled datasets. This innovative approach, employing twin Graph Neural Networks, has shown remarkable improvements in GNN performance across 14 benchmarks.

00:00

Playback language: English

Index

Abstract

Machine learning (ML) models have shown great promise in predicting material properties, but they often require large labeled datasets, which are expensive and time-consuming to generate. This paper introduces Crystal Twins (CT), a self-supervised learning (SSL) method that leverages large unlabeled datasets for crystalline material property prediction. CT employs a twin Graph Neural Network (GNN) architecture, learning representations by enforcing similarity between embeddings of augmented instances from the same crystalline system. Using Barlow Twins and SimSiam frameworks, CT significantly improves GNN performance on 14 material property prediction benchmarks.

Publisher

npj Computational Materials

Published On

Jan 31, 2022

Authors

Rishikesh Magar, Yuyang Wang, Amir Barati Farimani