Generative Adversarial Networks and Mixture Density Networks-Based Inverse Modeling for Microstructural Materials Design

Mao, Yuwei; Yang, Zijiang; Jha, Dipendra; Paul, Arindam; Liao, Wei-keng; Choudhary, Alok; Agrawal, Ankit

doi:10.1007/s40192-022-00285-0

Title: Generative Adversarial Networks and Mixture Density Networks-Based Inverse Modeling for Microstructural Materials Design

Abstract

Abstract There are two broad modeling paradigms in scientific applications: forward and inverse. While forward modeling estimates the observations based on known causes, inverse modeling attempts to infer the causes given the observations. Inverse problems are usually more critical as well as difficult in scientific applications as they seek to explore the causes that cannot be directly observed. Inverse problems are used extensively in various scientific fields, such as geophysics, health care and materials science. Exploring the relationships from properties to microstructures is one of the inverse problems in material science. It is challenging to solve the microstructure discovery inverse problem, because it usually needs to learn a one-to-many nonlinear mapping. Given a target property, there are multiple different microstructures that exhibit the target property, and their discovery also requires significant computing time. Further, microstructure discovery becomes even more difficult because the dimension of properties (input) is much lower than that of microstructures (output). In this work, we propose a framework consisting of generative adversarial networks and mixture density networks for inverse modeling of structure–property linkages in materials, i.e., microstructure discovery for a given property. The results demonstrate that compared to baseline methods, the proposed framework can overcome the above-mentionedmore »« less

Authors:

Mao, Yuwei; Yang, Zijiang; Jha, Dipendra; Paul, Arindam; Liao, Wei-keng; Choudhary, Alok;

Publication Date:: Tue Nov 08 00:00:00 EST 2022

Sponsoring Org.:: USDOE

OSTI Identifier:: 1897242

Grant/Contract Number:: SC0019358; SC0021399

Resource Type:: Published Article

Journal Name:: Integrating Materials and Manufacturing Innovation

Additional Journal Information:: Journal Name: Integrating Materials and Manufacturing Innovation Journal Volume: 11 Journal Issue: 4; Journal ID: ISSN 2193-9764

Publisher:: Springer Science + Business Media

Country of Publication:: Germany

Language:: English

Citation Formats


                    Mao, Yuwei, Yang, Zijiang, Jha, Dipendra, Paul, Arindam, Liao, Wei-keng, Choudhary, Alok, and Agrawal, Ankit. Generative Adversarial Networks and Mixture Density Networks-Based Inverse Modeling for Microstructural Materials Design.  Germany: N. p., 2022. 
Web.  doi:10.1007/s40192-022-00285-0.

Copy to clipboard


                    Mao, Yuwei, Yang, Zijiang, Jha, Dipendra, Paul, Arindam, Liao, Wei-keng, Choudhary, Alok, & Agrawal, Ankit. Generative Adversarial Networks and Mixture Density Networks-Based Inverse Modeling for Microstructural Materials Design.  Germany.  https://doi.org/10.1007/s40192-022-00285-0

Copy to clipboard


                    Mao, Yuwei, Yang, Zijiang, Jha, Dipendra, Paul, Arindam, Liao, Wei-keng, Choudhary, Alok, and Agrawal, Ankit. Tue .  
"Generative Adversarial Networks and Mixture Density Networks-Based Inverse Modeling for Microstructural Materials Design".  Germany.  https://doi.org/10.1007/s40192-022-00285-0.

Copy to clipboard


                    
@article{osti_1897242,

  title        = {Generative Adversarial Networks and Mixture Density Networks-Based Inverse Modeling for Microstructural Materials Design},

  author       = {Mao, Yuwei and Yang, Zijiang and Jha, Dipendra and Paul, Arindam and Liao, Wei-keng and Choudhary, Alok and Agrawal, Ankit},

  abstractNote = {Abstract There are two broad modeling paradigms in scientific applications: forward and inverse. While forward modeling estimates the observations based on known causes, inverse modeling attempts to infer the causes given the observations. Inverse problems are usually more critical as well as difficult in scientific applications as they seek to explore the causes that cannot be directly observed. Inverse problems are used extensively in various scientific fields, such as geophysics, health care and materials science. Exploring the relationships from properties to microstructures is one of the inverse problems in material science. It is challenging to solve the microstructure discovery inverse problem, because it usually needs to learn a one-to-many nonlinear mapping. Given a target property, there are multiple different microstructures that exhibit the target property, and their discovery also requires significant computing time. Further, microstructure discovery becomes even more difficult because the dimension of properties (input) is much lower than that of microstructures (output). In this work, we propose a framework consisting of generative adversarial networks and mixture density networks for inverse modeling of structure–property linkages in materials, i.e., microstructure discovery for a given property. The results demonstrate that compared to baseline methods, the proposed framework can overcome the above-mentioned challenges and discover multiple promising solutions in an efficient manner.},

  doi          = {10.1007/s40192-022-00285-0},

  journal      = {Integrating Materials and Manufacturing Innovation},

  number       = 4,

  volume       = 11,

  place        = {Germany},

  year         = {Tue Nov 08 00:00:00 EST 2022},

  month        = {Tue Nov 08 00:00:00 EST 2022}

}

Copy to clipboard

Journal Article:

Free Publicly Available Full Text

Publisher's Version of Record
https://doi.org/10.1007/s40192-022-00285-0

Other availability

Search WorldCat to find libraries that may hold this journal

Save / Share:

Export Metadata

Save to My Library

Works referenced in this record:

Perspective: Materials informatics and big data: Realization of the “fourth paradigm” of science in materials science
journal, April 2016

Agrawal, Ankit; Choudhary, Alok
APL Materials, Vol. 4, Issue 5
DOI: 10.1063/1.4946894

Deep learning for tomographic image reconstruction
journal, December 2020

Wang, Ge; Ye, Jong Chul; De Man, Bruno
Nature Machine Intelligence, Vol. 2, Issue 12
DOI: 10.1038/s42256-020-00273-z

U-Net: Convolutional Networks for Biomedical Image Segmentation
book, November 2015

Ronneberger, Olaf; Fischer, Philipp; Brox, Thomas
Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III
DOI: 10.1007/978-3-319-24574-4_28

Integrating physics-based models with sensor data: An inverse modeling approach
journal, May 2019

Hong, Tianzhen; Lee, Sang Hoon
Building and Environment, Vol. 154
DOI: 10.1016/j.buildenv.2019.03.006

Characterization and Design of Functional Quasi-Random Nanostructured Materials Using Spectral Density Function
journal, May 2017

Yu, Shuangcheng; Zhang, Yichi; Wang, Chen
Journal of Mechanical Design, Vol. 139, Issue 7
DOI: 10.1115/1.4036582

Computational materials science and engineering education: A survey of trends and needs
journal, October 2009

Thornton, K.; Nola, Samanthule; Edwin Garcia, R.
JOM, Vol. 61, Issue 10
DOI: 10.1007/s11837-009-0142-3

Deep Residual Learning for Model-Based Iterative CT Reconstruction Using Plug-and-Play Framework
conference, April 2018

Ye, Dong Hye; Srivastava, Somesh; Thibault, Jean-Baptiste
ICASSP 2018 - 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
DOI: 10.1109/ICASSP.2018.8461408

Densely Connected Convolutional Networks
conference, July 2017

Huang, Gao; Liu, Zhuang; Maaten, Laurens van der
2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
DOI: 10.1109/CVPR.2017.243

Context Aware Machine Learning Approaches for Modeling Elastic Localization in Three-Dimensional Composite Microstructures
journal, March 2017

Liu, Ruoqian; Yabansu, Yuksel C.; Yang, Zijiang
Integrating Materials and Manufacturing Innovation, Vol. 6, Issue 2
DOI: 10.1007/s40192-017-0094-3

A Deep Adversarial Learning Methodology for Designing Microstructural Material Systems
conference, November 2018

Li, Xiaolin; Yang, Zijiang; Brinson, L. Catherine
ASME 2018 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, Volume 2B: 44th Design Automation Conference
DOI: 10.1115/DETC2018-85633

Accelerated materials property predictions and design using motif-based fingerprints
journal, July 2015

Huan, Tran Doan; Mannodi-Kanakkithodi, Arun; Ramprasad, Rampi
Physical Review B, Vol. 92, Issue 1
DOI: 10.1103/PhysRevB.92.014106

Cross-property deep transfer learning framework for enhanced predictive analytics on small materials data
journal, November 2021

Gupta, Vishu; Choudhary, Kamal; Tavazza, Francesca
Nature Communications, Vol. 12, Issue 1
DOI: 10.1038/s41467-021-26921-5

Inverse modeling for quantitative X-ray microanalysis applied to 2D heterogeneous materials
journal, December 2020

Yuan, Yu; Demers, Hendrix; Brodusch, Nicolas
Ultramicroscopy, Vol. 219
DOI: 10.1016/j.ultramic.2020.113117

Deep learning approaches for mining structure-property linkages in high contrast composites from simulation datasets
journal, August 2018

Yang, Zijiang; Yabansu, Yuksel C.; Al-Bahrani, Reda
Computational Materials Science, Vol. 151
DOI: 10.1016/j.commatsci.2018.05.014

Data-Driven Multi-Scale Modeling and Optimization for Elastic Properties of Cubic Microstructures
journal, April 2022

Hasan, M.; Mao, Y.; Choudhary, K.
Integrating Materials and Manufacturing Innovation, Vol. 11, Issue 2
DOI: 10.1007/s40192-022-00258-3

An integrated inverse model-experimental approach to determine soft tissue three-dimensional constitutive parameters: application to post-infarcted myocardium
journal, August 2017

Avazmohammadi, Reza; Li, David S.; Leahy, Thomas
Biomechanics and Modeling in Mechanobiology, Vol. 17, Issue 1
DOI: 10.1007/s10237-017-0943-1

StackGAN: Text to Photo-Realistic Image Synthesis with Stacked Generative Adversarial Networks
conference, October 2017

Zhang, Han; Xu, Tao; Li, Hongsheng
2017 IEEE International Conference on Computer Vision (ICCV)
DOI: 10.1109/ICCV.2017.629

EBSD image segmentation using a physics-based forward model
conference, September 2013

Park, Se Un; Wei, Dennis; De Graef, Marc
2013 20th IEEE International Conference on Image Processing (ICIP), 2013 IEEE International Conference on Image Processing
DOI: 10.1109/ICIP.2013.6738779

A Novel Method for Determination of Dielectric Properties of Materials Using a Combined Embedded Modulated Scattering and Near-Field Microwave Techniques—Part I: Forward Model
journal, December 2005

Hughes, D.; Zoughi, R.
IEEE Transactions on Instrumentation and Measurement, Vol. 54, Issue 6
DOI: 10.1109/TIM.2005.858132

Microstructural Materials Design Via Deep Adversarial Learning Methodology
journal, October 2018

Yang, Zijiang; Li, Xiaolin; Catherine Brinson, L.
Journal of Mechanical Design, Vol. 140, Issue 11
DOI: 10.1115/1.4041371

Deep materials informatics: Applications of deep learning in materials science
journal, June 2019

Agrawal, Ankit; Choudhary, Alok
MRS Communications, Vol. 9, Issue 3
DOI: 10.1557/mrc.2019.73

BRNet: Branched Residual Network for Fast and Accurate Predictive Modeling of Materials Properties
book, January 2022

Gupta, Vishu; Liao, Wei-keng; Choudhary, Alok
Proceedings of the 2022 SIAM International Conference on Data Mining (SDM)
DOI: 10.1137/1.9781611977172.39

Deep Recurrent Models with Fast-Forward Connections for Neural Machine Translation
journal, December 2016

Zhou, Jie; Cao, Ying; Wang, Xuguang
Transactions of the Association for Computational Linguistics, Vol. 4
DOI: 10.1162/tacl_a_00105

Establishing structure-property localization linkages for elastic deformation of three-dimensional high contrast composites using deep learning approaches
journal, March 2019

Yang, Zijiang; Yabansu, Yuksel C.; Jha, Dipendra
Acta Materialia, Vol. 166
DOI: 10.1016/j.actamat.2018.12.045

The joint automated repository for various integrated simulations (JARVIS) for data-driven materials design
journal, November 2020

Choudhary, Kamal; Garrity, Kevin F.; Reid, Andrew C. E.
npj Computational Materials, Vol. 6, Issue 1
DOI: 10.1038/s41524-020-00440-1

Deep Residual Learning for Image Recognition
conference, June 2016

He, Kaiming; Zhang, Xiangyu; Ren, Shaoqing
2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
DOI: 10.1109/CVPR.2016.90

Min-Entropy Latent Model for Weakly Supervised Object Detection
conference, June 2018

Wan, Fang; Wei, Pengxu; Jiao, Jianbin
2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition
DOI: 10.1109/CVPR.2018.00141

Mathematical and experimental method to obtain the inverse modeling of nonsinusoidal and saturated synchronous reluctance motors
journal, December 2003

Sturtzer, G.; Flieller, D.; Louis, J. P.
IEEE Transactions on Energy Conversion, Vol. 18, Issue 4
DOI: 10.1109/TEC.2003.816601

Inverse Modeling Applied for Material Characterization of Powder Materials
journal, October 2014

Andersson, Daniel C.; Lindskog, Per; Larsson, Per-Lennart
Journal of Testing and Evaluation, Vol. 43, Issue 5
DOI: 10.1520/JTE20130266

Enabling deeper learning on big data for materials informatics applications
journal, February 2021

Jha, Dipendra; Gupta, Vishu; Ward, Logan
Scientific Reports, Vol. 11, Issue 1
DOI: 10.1038/s41598-021-83193-1

Combinatorial screening for new materials in unconstrained composition space with machine learning
journal, March 2014

Meredig, B.; Agrawal, A.; Kirklin, S.
Physical Review B, Vol. 89, Issue 9
DOI: 10.1103/PhysRevB.89.094104

Similar Records in DOE PAGES and OSTI.GOV collections:

Simultaneously improving accuracy and computational cost under parametric constraints in materials property prediction tasks

Journal Article Gupta, Vishu ; Li, Youjia ; Peltekian, Alec ; ... - Journal of Cheminformatics

Abstract Modern data mining techniques using machine learning (ML) and deep learning (DL) algorithms have been shown to excel in the regression-based task of materials property prediction using various materials representations. In an attempt to improve the predictive performance of the deep neural network model, researchers have tried to add more layers as well as develop new architectural components to create sophisticated and deep neural network models that can aid in the training process and improve the predictive ability of the final model. However, usually, these modifications require a lot of computational resources, thereby further increasing the already large modelmore »« less
https://doi.org/10.1186/s13321-024-00811-6
Structure-aware graph neural network based deep transfer learning framework for enhanced predictive analytics on diverse materials datasets

Journal Article Gupta, Vishu ; Choudhary, Kamal ; DeCost, Brian ; ... - npj Computational Materials

Abstract Modern data mining methods have demonstrated effectiveness in comprehending and predicting materials properties. An essential component in the process of materials discovery is to know which material(s) will possess desirable properties. For many materials properties, performing experiments and density functional theory computations are costly and time-consuming. Hence, it is challenging to build accurate predictive models for such properties using conventional data mining methods due to the small amount of available data. Here we present a framework for materials property prediction tasks using structure information that leverages graph neural network-based architecture along with deep-transfer-learning techniques to drastically improve the model’smore »« less
https://doi.org/10.1038/s41524-023-01185-3
Evolution of artificial intelligence for application in contemporary materials science

Journal Article Gupta, Vishu ; Liao, Wei-keng ; Choudhary, Alok ; ... - MRS communications

Abstract Contemporary materials science has seen an increasing application of various artificial intelligence techniques in an attempt to accelerate the materials discovery process using forward modeling for predictive analysis and inverse modeling for optimization and design. Over the last decade or so, the increasing availability of computational power and large materials datasets has led to a continuous evolution in the complexity of the techniques used to advance the frontier. In this Review, we provide a high-level overview of the evolution of artificial intelligence in contemporary materials science for the task of materials property prediction in forward modeling. Each stage ofmore »« less
https://doi.org/10.1557/s43579-023-00433-3
Challenges at the Frontiers of Matter and Energy: Transformative Opportunities for Discovery Science

Program Document Hemminger, John C. ; Sarrao, John ; Crabtree, George ; ...

FIVE TRANSFORMATIVE OPPORTUNITIES FOR DISCOVERY SCIENCE As a result of this effort, it has become clear that the progress made to date on the five Grand Challenges has created a springboard for seizing five new Transformative Opportunities that have the potential to further transform key technologies involving matter and energy. These five new Transformative Opportunities and the evidence supporting them are discussed in this new report, “Challenges at the Frontiers of Matter and Energy: Transformative Opportunities for Discovery Science.” Mastering Hierarchical Architectures and Beyond-Equilibrium Matter Complex materials and chemical processes transmute matter and energy, for example from CO2 and watermore »« less
https://doi.org/10.2172/1283188

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

Journal Article Gupta, Vishu ; Choudhary, Kamal ; Tavazza, Francesca ; ... - Nature Communications

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 fractionsmore »« less
https://doi.org/10.1038/s41467-021-26921-5

Similar Records

Title: Generative Adversarial Networks and Mixture Density Networks-Based Inverse Modeling for Microstructural Materials Design

Abstract

Citation Formats

Perspective: Materials informatics and big data: Realization of the “fourth paradigm” of science in materials science journal, April 2016

Deep learning for tomographic image reconstruction journal, December 2020

U-Net: Convolutional Networks for Biomedical Image Segmentation book, November 2015

Integrating physics-based models with sensor data: An inverse modeling approach journal, May 2019

Characterization and Design of Functional Quasi-Random Nanostructured Materials Using Spectral Density Function journal, May 2017

Computational materials science and engineering education: A survey of trends and needs journal, October 2009

Deep Residual Learning for Model-Based Iterative CT Reconstruction Using Plug-and-Play Framework conference, April 2018

Densely Connected Convolutional Networks conference, July 2017

Context Aware Machine Learning Approaches for Modeling Elastic Localization in Three-Dimensional Composite Microstructures journal, March 2017

A Deep Adversarial Learning Methodology for Designing Microstructural Material Systems conference, November 2018

Accelerated materials property predictions and design using motif-based fingerprints journal, July 2015

Cross-property deep transfer learning framework for enhanced predictive analytics on small materials data journal, November 2021

Inverse modeling for quantitative X-ray microanalysis applied to 2D heterogeneous materials journal, December 2020

Deep learning approaches for mining structure-property linkages in high contrast composites from simulation datasets journal, August 2018

Data-Driven Multi-Scale Modeling and Optimization for Elastic Properties of Cubic Microstructures journal, April 2022

An integrated inverse model-experimental approach to determine soft tissue three-dimensional constitutive parameters: application to post-infarcted myocardium journal, August 2017

StackGAN: Text to Photo-Realistic Image Synthesis with Stacked Generative Adversarial Networks conference, October 2017

EBSD image segmentation using a physics-based forward model conference, September 2013

A Novel Method for Determination of Dielectric Properties of Materials Using a Combined Embedded Modulated Scattering and Near-Field Microwave Techniques—Part I: Forward Model journal, December 2005

Microstructural Materials Design Via Deep Adversarial Learning Methodology journal, October 2018

Deep materials informatics: Applications of deep learning in materials science journal, June 2019

BRNet: Branched Residual Network for Fast and Accurate Predictive Modeling of Materials Properties book, January 2022

Deep Recurrent Models with Fast-Forward Connections for Neural Machine Translation journal, December 2016

Establishing structure-property localization linkages for elastic deformation of three-dimensional high contrast composites using deep learning approaches journal, March 2019

The joint automated repository for various integrated simulations (JARVIS) for data-driven materials design journal, November 2020

Deep Residual Learning for Image Recognition conference, June 2016

Min-Entropy Latent Model for Weakly Supervised Object Detection conference, June 2018

Mathematical and experimental method to obtain the inverse modeling of nonsinusoidal and saturated synchronous reluctance motors journal, December 2003

Inverse Modeling Applied for Material Characterization of Powder Materials journal, October 2014

Enabling deeper learning on big data for materials informatics applications journal, February 2021

Combinatorial screening for new materials in unconstrained composition space with machine learning journal, March 2014

Perspective: Materials informatics and big data: Realization of the “fourth paradigm” of science in materials science
journal, April 2016

Deep learning for tomographic image reconstruction
journal, December 2020

U-Net: Convolutional Networks for Biomedical Image Segmentation
book, November 2015

Integrating physics-based models with sensor data: An inverse modeling approach
journal, May 2019

Characterization and Design of Functional Quasi-Random Nanostructured Materials Using Spectral Density Function
journal, May 2017

Computational materials science and engineering education: A survey of trends and needs
journal, October 2009

Deep Residual Learning for Model-Based Iterative CT Reconstruction Using Plug-and-Play Framework
conference, April 2018

Densely Connected Convolutional Networks
conference, July 2017

Context Aware Machine Learning Approaches for Modeling Elastic Localization in Three-Dimensional Composite Microstructures
journal, March 2017

A Deep Adversarial Learning Methodology for Designing Microstructural Material Systems
conference, November 2018

Accelerated materials property predictions and design using motif-based fingerprints
journal, July 2015

Cross-property deep transfer learning framework for enhanced predictive analytics on small materials data
journal, November 2021

Inverse modeling for quantitative X-ray microanalysis applied to 2D heterogeneous materials
journal, December 2020

Deep learning approaches for mining structure-property linkages in high contrast composites from simulation datasets
journal, August 2018

Data-Driven Multi-Scale Modeling and Optimization for Elastic Properties of Cubic Microstructures
journal, April 2022

An integrated inverse model-experimental approach to determine soft tissue three-dimensional constitutive parameters: application to post-infarcted myocardium
journal, August 2017

StackGAN: Text to Photo-Realistic Image Synthesis with Stacked Generative Adversarial Networks
conference, October 2017

EBSD image segmentation using a physics-based forward model
conference, September 2013

A Novel Method for Determination of Dielectric Properties of Materials Using a Combined Embedded Modulated Scattering and Near-Field Microwave Techniques—Part I: Forward Model
journal, December 2005

Microstructural Materials Design Via Deep Adversarial Learning Methodology
journal, October 2018

Deep materials informatics: Applications of deep learning in materials science
journal, June 2019

BRNet: Branched Residual Network for Fast and Accurate Predictive Modeling of Materials Properties
book, January 2022

Deep Recurrent Models with Fast-Forward Connections for Neural Machine Translation
journal, December 2016

Establishing structure-property localization linkages for elastic deformation of three-dimensional high contrast composites using deep learning approaches
journal, March 2019

The joint automated repository for various integrated simulations (JARVIS) for data-driven materials design
journal, November 2020

Deep Residual Learning for Image Recognition
conference, June 2016

Min-Entropy Latent Model for Weakly Supervised Object Detection
conference, June 2018

Mathematical and experimental method to obtain the inverse modeling of nonsinusoidal and saturated synchronous reluctance motors
journal, December 2003

Inverse Modeling Applied for Material Characterization of Powder Materials
journal, October 2014

Enabling deeper learning on big data for materials informatics applications
journal, February 2021

Combinatorial screening for new materials in unconstrained composition space with machine learning
journal, March 2014