Interpreting convolutional neural network classifiers applied to laser-induced breakdown optical emission spectra

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26620%2F24%3APU155235" target="_blank" >RIV/00216305:26620/24:PU155235 - isvavai.cz</a>

Nalezeny alternativní kódy

RIV/00216224:14330/24:00137458

Výsledek na webu

<a href="https://www-sciencedirect-com.ezproxy.lib.vutbr.cz/science/article/pii/S0039914023006975?via%3Dihub" target="_blank" >https://www-sciencedirect-com.ezproxy.lib.vutbr.cz/science/article/pii/S0039914023006975?via%3Dihub</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.talanta.2023.124946" target="_blank" >10.1016/j.talanta.2023.124946</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Interpreting convolutional neural network classifiers applied to laser-induced breakdown optical emission spectra

Popis výsledku v původním jazyce

Laser-induced breakdown spectroscopy (LIBS) is a well-established industrial tool with emerging relevance in high-stakes applications. To achieve its required analytical performance, LIBS is often coupled with advanced pattern-recognition algorithms, including machine learning models. Namely, artificial neural networks (ANNs) have recently become a frequently applied part of LIBS practitioners' toolkit. Nevertheless, ANNs are generally applied in spectroscopy as black-box models, without a real insight into their predictions. Here, we apply various post-hoc interpretation techniques with the aim of understanding the decision-making of convolutional neural networks. Namely, we find synthetic spectra that yield perfect expected classification predictions and denote these spectra class-specific prototype spectra. We investigate the simplest possible convolutional neural network (consisting of a single convolutional and fully connected layers) trained to classify the extended calibration dataset collected for the ChemCam laser-induced breakdown spectroscopy instrument of the Curiosity Mars rover. The trained convolutional neural network predominantly learned meaningful spectroscopic features which correspond to the elements comprising the major oxides found in the calibration targets. In addition, the discrete convolution operation with the learnt filters results in a crude baseline correction.

Název v anglickém jazyce

Interpreting convolutional neural network classifiers applied to laser-induced breakdown optical emission spectra

Popis výsledku anglicky

Laser-induced breakdown spectroscopy (LIBS) is a well-established industrial tool with emerging relevance in high-stakes applications. To achieve its required analytical performance, LIBS is often coupled with advanced pattern-recognition algorithms, including machine learning models. Namely, artificial neural networks (ANNs) have recently become a frequently applied part of LIBS practitioners' toolkit. Nevertheless, ANNs are generally applied in spectroscopy as black-box models, without a real insight into their predictions. Here, we apply various post-hoc interpretation techniques with the aim of understanding the decision-making of convolutional neural networks. Namely, we find synthetic spectra that yield perfect expected classification predictions and denote these spectra class-specific prototype spectra. We investigate the simplest possible convolutional neural network (consisting of a single convolutional and fully connected layers) trained to classify the extended calibration dataset collected for the ChemCam laser-induced breakdown spectroscopy instrument of the Curiosity Mars rover. The trained convolutional neural network predominantly learned meaningful spectroscopic features which correspond to the elements comprising the major oxides found in the calibration targets. In addition, the discrete convolution operation with the learnt filters results in a crude baseline correction.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10406 - Analytical chemistry

Projekt

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2024

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

TALANTA

ISSN

0039-9140

e-ISSN

1873-3573

Svazek periodika

266

Číslo periodika v rámci svazku

1

Stát vydavatele periodika

GB - Spojené království Velké Británie a Severního Irska

Počet stran výsledku

11

Strana od-do

„“-„“

Kód UT WoS článku

001044992600001

EID výsledku v databázi Scopus

2-s2.0-85165081844