A study of real-world micrograph data quality and machine learning model robustness

Abstract Machine-learning (ML) techniques hold the potential of enabling efficient quantitative micrograph analysis, but the robustness of ML models with respect to real-world micrograph quality variations has not been carefully evaluated. We collected thousands of scanning electron microscopy (SEM)...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Xiaoting Zhong, Brian Gallagher, Keenan Eves, Emily Robertson, T. Nathan Mundhenk, T. Yong-Jin Han
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Materials of engineering and construction. Mechanics of materials TA401-492 Computer software QA76.75-76.765
Acceso en línea:	https://doaj.org/article/15813301f7d54bfcb75f7018545b366d
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:15813301f7d54bfcb75f7018545b366d
record_format	dspace
spelling	oai:doaj.org-article:15813301f7d54bfcb75f7018545b366d2021-12-02T19:16:18ZA study of real-world micrograph data quality and machine learning model robustness10.1038/s41524-021-00616-32057-3960https://doaj.org/article/15813301f7d54bfcb75f7018545b366d2021-10-01T00:00:00Zhttps://doi.org/10.1038/s41524-021-00616-3https://doaj.org/toc/2057-3960Abstract Machine-learning (ML) techniques hold the potential of enabling efficient quantitative micrograph analysis, but the robustness of ML models with respect to real-world micrograph quality variations has not been carefully evaluated. We collected thousands of scanning electron microscopy (SEM) micrographs for molecular solid materials, in which image pixel intensities vary due to both the microstructure content and microscope instrument conditions. We then built ML models to predict the ultimate compressive strength (UCS) of consolidated molecular solids, by encoding micrographs with different image feature descriptors and training a random forest regressor, and by training an end-to-end deep-learning (DL) model. Results show that instrument-induced pixel intensity signals can affect ML model predictions in a consistently negative way. As a remedy, we explored intensity normalization techniques. It is seen that intensity normalization helps to improve micrograph data quality and ML model robustness, but microscope-induced intensity variations can be difficult to eliminate.Xiaoting ZhongBrian GallagherKeenan EvesEmily RobertsonT. Nathan MundhenkT. Yong-Jin HanNature PortfolioarticleMaterials of engineering and construction. Mechanics of materialsTA401-492Computer softwareQA76.75-76.765ENnpj Computational Materials, Vol 7, Iss 1, Pp 1-11 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Materials of engineering and construction. Mechanics of materials TA401-492 Computer software QA76.75-76.765
spellingShingle	Materials of engineering and construction. Mechanics of materials TA401-492 Computer software QA76.75-76.765 Xiaoting Zhong Brian Gallagher Keenan Eves Emily Robertson T. Nathan Mundhenk T. Yong-Jin Han A study of real-world micrograph data quality and machine learning model robustness
description	Abstract Machine-learning (ML) techniques hold the potential of enabling efficient quantitative micrograph analysis, but the robustness of ML models with respect to real-world micrograph quality variations has not been carefully evaluated. We collected thousands of scanning electron microscopy (SEM) micrographs for molecular solid materials, in which image pixel intensities vary due to both the microstructure content and microscope instrument conditions. We then built ML models to predict the ultimate compressive strength (UCS) of consolidated molecular solids, by encoding micrographs with different image feature descriptors and training a random forest regressor, and by training an end-to-end deep-learning (DL) model. Results show that instrument-induced pixel intensity signals can affect ML model predictions in a consistently negative way. As a remedy, we explored intensity normalization techniques. It is seen that intensity normalization helps to improve micrograph data quality and ML model robustness, but microscope-induced intensity variations can be difficult to eliminate.
format	article
author	Xiaoting Zhong Brian Gallagher Keenan Eves Emily Robertson T. Nathan Mundhenk T. Yong-Jin Han
author_facet	Xiaoting Zhong Brian Gallagher Keenan Eves Emily Robertson T. Nathan Mundhenk T. Yong-Jin Han
author_sort	Xiaoting Zhong
title	A study of real-world micrograph data quality and machine learning model robustness
title_short	A study of real-world micrograph data quality and machine learning model robustness
title_full	A study of real-world micrograph data quality and machine learning model robustness
title_fullStr	A study of real-world micrograph data quality and machine learning model robustness
title_full_unstemmed	A study of real-world micrograph data quality and machine learning model robustness
title_sort	study of real-world micrograph data quality and machine learning model robustness
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/15813301f7d54bfcb75f7018545b366d
work_keys_str_mv	AT xiaotingzhong astudyofrealworldmicrographdataqualityandmachinelearningmodelrobustness AT briangallagher astudyofrealworldmicrographdataqualityandmachinelearningmodelrobustness AT keenaneves astudyofrealworldmicrographdataqualityandmachinelearningmodelrobustness AT emilyrobertson astudyofrealworldmicrographdataqualityandmachinelearningmodelrobustness AT tnathanmundhenk astudyofrealworldmicrographdataqualityandmachinelearningmodelrobustness AT tyongjinhan astudyofrealworldmicrographdataqualityandmachinelearningmodelrobustness AT xiaotingzhong studyofrealworldmicrographdataqualityandmachinelearningmodelrobustness AT briangallagher studyofrealworldmicrographdataqualityandmachinelearningmodelrobustness AT keenaneves studyofrealworldmicrographdataqualityandmachinelearningmodelrobustness AT emilyrobertson studyofrealworldmicrographdataqualityandmachinelearningmodelrobustness AT tnathanmundhenk studyofrealworldmicrographdataqualityandmachinelearningmodelrobustness AT tyongjinhan studyofrealworldmicrographdataqualityandmachinelearningmodelrobustness
_version_	1718376957120348160

A study of real-world micrograph data quality and machine learning model robustness

Ejemplares similares