Machine learning for pattern and waveform recognitions in terahertz image data

Abstract Several machine learning (ML) techniques were tested for the feasibility of performing automated pattern and waveform recognitions of terahertz time-domain spectroscopy datasets. Out of all the ML techniques under test, it was observed that random forest statistical algorithm works well wit...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Dmitry S. Bulgarevich, Miezel Talara, Masahiko Tani, Makoto Watanabe
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/20abf015762a401f86ff26e42bd44bed
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:20abf015762a401f86ff26e42bd44bed
record_format	dspace
spelling	oai:doaj.org-article:20abf015762a401f86ff26e42bd44bed2021-12-02T14:01:23ZMachine learning for pattern and waveform recognitions in terahertz image data10.1038/s41598-020-80761-92045-2322https://doaj.org/article/20abf015762a401f86ff26e42bd44bed2021-01-01T00:00:00Zhttps://doi.org/10.1038/s41598-020-80761-9https://doaj.org/toc/2045-2322Abstract Several machine learning (ML) techniques were tested for the feasibility of performing automated pattern and waveform recognitions of terahertz time-domain spectroscopy datasets. Out of all the ML techniques under test, it was observed that random forest statistical algorithm works well with the THz datasets in both the frequency and time domains. With such ML algorithm, a classifier can be created with less than 1% out-of-bag error for segmentation of rusted and non-rusted sample regions of the image datasets in frequency domain. The degree of linear correlation between the rusted area percentage and the image spatial resolution with terahertz frequency can be used as an additional cross-validation criteria for the evaluation of classifier quality. However, for different rust staging measured datasets, a standardized procedure of image pre-processing is necessary to create/apply a single classifier and its usage is only limited to 1 ± 0.2 THz. Moreover, random forest is practically the best choice among the several popular ML techniques under test for waveform recognition of time-domain data in terms of classification accuracy and timing. Our results demonstrate the usefulness of random forest and several other machine learning algorithms for terahertz hyperspectral pattern recognition.Dmitry S. BulgarevichMiezel TalaraMasahiko TaniMakoto WatanabeNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-8 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Dmitry S. Bulgarevich Miezel Talara Masahiko Tani Makoto Watanabe Machine learning for pattern and waveform recognitions in terahertz image data
description	Abstract Several machine learning (ML) techniques were tested for the feasibility of performing automated pattern and waveform recognitions of terahertz time-domain spectroscopy datasets. Out of all the ML techniques under test, it was observed that random forest statistical algorithm works well with the THz datasets in both the frequency and time domains. With such ML algorithm, a classifier can be created with less than 1% out-of-bag error for segmentation of rusted and non-rusted sample regions of the image datasets in frequency domain. The degree of linear correlation between the rusted area percentage and the image spatial resolution with terahertz frequency can be used as an additional cross-validation criteria for the evaluation of classifier quality. However, for different rust staging measured datasets, a standardized procedure of image pre-processing is necessary to create/apply a single classifier and its usage is only limited to 1 ± 0.2 THz. Moreover, random forest is practically the best choice among the several popular ML techniques under test for waveform recognition of time-domain data in terms of classification accuracy and timing. Our results demonstrate the usefulness of random forest and several other machine learning algorithms for terahertz hyperspectral pattern recognition.
format	article
author	Dmitry S. Bulgarevich Miezel Talara Masahiko Tani Makoto Watanabe
author_facet	Dmitry S. Bulgarevich Miezel Talara Masahiko Tani Makoto Watanabe
author_sort	Dmitry S. Bulgarevich
title	Machine learning for pattern and waveform recognitions in terahertz image data
title_short	Machine learning for pattern and waveform recognitions in terahertz image data
title_full	Machine learning for pattern and waveform recognitions in terahertz image data
title_fullStr	Machine learning for pattern and waveform recognitions in terahertz image data
title_full_unstemmed	Machine learning for pattern and waveform recognitions in terahertz image data
title_sort	machine learning for pattern and waveform recognitions in terahertz image data
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/20abf015762a401f86ff26e42bd44bed
work_keys_str_mv	AT dmitrysbulgarevich machinelearningforpatternandwaveformrecognitionsinterahertzimagedata AT miezeltalara machinelearningforpatternandwaveformrecognitionsinterahertzimagedata AT masahikotani machinelearningforpatternandwaveformrecognitionsinterahertzimagedata AT makotowatanabe machinelearningforpatternandwaveformrecognitionsinterahertzimagedata
_version_	1718392126952177664

Machine learning for pattern and waveform recognitions in terahertz image data

Ejemplares similares