XElemNet: towards explainable AI for deep neural networks in materials science

Abstract Recent progress in deep learning has significantly impacted materials science, leading to accelerated material discovery and innovation. ElemNet, a deep neural network model that predicts formation energy from elemental compositions, exemplifies the application of deep learning techniques i...

Full description

Saved in:
Bibliographic Details
Main Authors: Kewei Wang, Vishu Gupta, Claire Songhyun Lee, Yuwei Mao, Muhammed Nur Talha Kilic, Youjia Li, Zanhua Huang, Wei-keng Liao, Alok Choudhary, Ankit Agrawal
Format: Article
Language:English
Published: Nature Portfolio 2024-10-01
Series:Scientific Reports
Online Access:https://doi.org/10.1038/s41598-024-76535-2
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850203875335733248
author Kewei Wang
Vishu Gupta
Claire Songhyun Lee
Yuwei Mao
Muhammed Nur Talha Kilic
Youjia Li
Zanhua Huang
Wei-keng Liao
Alok Choudhary
Ankit Agrawal
author_facet Kewei Wang
Vishu Gupta
Claire Songhyun Lee
Yuwei Mao
Muhammed Nur Talha Kilic
Youjia Li
Zanhua Huang
Wei-keng Liao
Alok Choudhary
Ankit Agrawal
author_sort Kewei Wang
collection DOAJ
description Abstract Recent progress in deep learning has significantly impacted materials science, leading to accelerated material discovery and innovation. ElemNet, a deep neural network model that predicts formation energy from elemental compositions, exemplifies the application of deep learning techniques in this field. However, the “black-box” nature of deep learning models often raises concerns about their interpretability and reliability. In this study, we propose XElemNet to explore the interpretability of ElemNet by applying a series of explainable artificial intelligence (XAI) techniques, focusing on post-hoc analysis and model transparency. The experiments with artificial binary datasets reveal ElemNet’s effectiveness in predicting convex hulls of element-pair systems across periodic table groups, indicating its capability to effectively discern elemental interactions in most cases. Additionally, feature importance analysis within ElemNet highlights alignment with chemical properties of elements such as reactivity and electronegativity. XElemNet provides insights into the strengths and limitations of ElemNet and offers a potential pathway for explaining other deep learning models in materials science.
format Article
id doaj-art-0b3ce6ce58b2483e894f5351995d1ca5
institution OA Journals
issn 2045-2322
language English
publishDate 2024-10-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-0b3ce6ce58b2483e894f5351995d1ca52025-08-20T02:11:25ZengNature PortfolioScientific Reports2045-23222024-10-0114111210.1038/s41598-024-76535-2XElemNet: towards explainable AI for deep neural networks in materials scienceKewei Wang0Vishu Gupta1Claire Songhyun Lee2Yuwei Mao3Muhammed Nur Talha Kilic4Youjia Li5Zanhua Huang6Wei-keng Liao7Alok Choudhary8Ankit Agrawal9Electrical and Computer Engineering, Northwestern UniversityElectrical and Computer Engineering, Northwestern UniversityElectrical and Computer Engineering, Northwestern UniversityElectrical and Computer Engineering, Northwestern UniversityElectrical and Computer Engineering, Northwestern UniversityElectrical and Computer Engineering, Northwestern UniversityElectrical and Computer Engineering, Northwestern UniversityElectrical and Computer Engineering, Northwestern UniversityElectrical and Computer Engineering, Northwestern UniversityElectrical and Computer Engineering, Northwestern UniversityAbstract Recent progress in deep learning has significantly impacted materials science, leading to accelerated material discovery and innovation. ElemNet, a deep neural network model that predicts formation energy from elemental compositions, exemplifies the application of deep learning techniques in this field. However, the “black-box” nature of deep learning models often raises concerns about their interpretability and reliability. In this study, we propose XElemNet to explore the interpretability of ElemNet by applying a series of explainable artificial intelligence (XAI) techniques, focusing on post-hoc analysis and model transparency. The experiments with artificial binary datasets reveal ElemNet’s effectiveness in predicting convex hulls of element-pair systems across periodic table groups, indicating its capability to effectively discern elemental interactions in most cases. Additionally, feature importance analysis within ElemNet highlights alignment with chemical properties of elements such as reactivity and electronegativity. XElemNet provides insights into the strengths and limitations of ElemNet and offers a potential pathway for explaining other deep learning models in materials science.https://doi.org/10.1038/s41598-024-76535-2
spellingShingle Kewei Wang
Vishu Gupta
Claire Songhyun Lee
Yuwei Mao
Muhammed Nur Talha Kilic
Youjia Li
Zanhua Huang
Wei-keng Liao
Alok Choudhary
Ankit Agrawal
XElemNet: towards explainable AI for deep neural networks in materials science
Scientific Reports
title XElemNet: towards explainable AI for deep neural networks in materials science
title_full XElemNet: towards explainable AI for deep neural networks in materials science
title_fullStr XElemNet: towards explainable AI for deep neural networks in materials science
title_full_unstemmed XElemNet: towards explainable AI for deep neural networks in materials science
title_short XElemNet: towards explainable AI for deep neural networks in materials science
title_sort xelemnet towards explainable ai for deep neural networks in materials science
url https://doi.org/10.1038/s41598-024-76535-2
work_keys_str_mv AT keweiwang xelemnettowardsexplainableaifordeepneuralnetworksinmaterialsscience
AT vishugupta xelemnettowardsexplainableaifordeepneuralnetworksinmaterialsscience
AT clairesonghyunlee xelemnettowardsexplainableaifordeepneuralnetworksinmaterialsscience
AT yuweimao xelemnettowardsexplainableaifordeepneuralnetworksinmaterialsscience
AT muhammednurtalhakilic xelemnettowardsexplainableaifordeepneuralnetworksinmaterialsscience
AT youjiali xelemnettowardsexplainableaifordeepneuralnetworksinmaterialsscience
AT zanhuahuang xelemnettowardsexplainableaifordeepneuralnetworksinmaterialsscience
AT weikengliao xelemnettowardsexplainableaifordeepneuralnetworksinmaterialsscience
AT alokchoudhary xelemnettowardsexplainableaifordeepneuralnetworksinmaterialsscience
AT ankitagrawal xelemnettowardsexplainableaifordeepneuralnetworksinmaterialsscience