Detail publikace

Interpreting neural networks trained to predict plasma temperature from optical emission spectra

KÉPEŠ, E. SAEIDFIROUZEH, H. LAITL, V. VRÁBEL, J. KUBELÍK, P. POŘÍZKA, P. FERUS, M. KAISER, J.

Originální název

Interpreting neural networks trained to predict plasma temperature from optical emission spectra

Typ

článek v časopise ve Web of Science, Jimp

Jazyk

angličtina

Originální abstrakt

We explore the application of artificial neural networks (ANNs) for predicting plasma temperatures in Laser-Induced Breakdown Spectroscopy (LIBS) analysis. Estimating plasma temperature from emission spectra is often challenging due to spectral interference and matrix effects. Traditional methods like the Boltzmann plot technique have limitations, both in applicability due to various matrix effects and in accuracy owing to the uncertainty of the underlying spectroscopic constants. Consequently, ANNs have already been successfully demonstrated as a viable alternative for plasma temperature prediction. We leverage synthetic data to isolate temperature effects from other factors and study the relationship between the LIBS spectra and temperature learnt by the ANN. We employ various post-hoc model interpretation techniques, including gradient-based methods, to verify that ANNs learn meaningful spectroscopic features for temperature prediction. Our findings demonstrate the potential of ANNs to learn complex relationships in LIBS spectra, offering a promising avenue for improved plasma temperature estimation and enhancing the overall accuracy of LIBS analysis. ANN can learn spectroscopic trends widely used by domain experts for plasma temperature estimation using emission spectra.

Klíčová slova

INDUCED BREAKDOWN SPECTROSCOPY; LASER-INDUCED PLASMA; CHEMCAM INSTRUMENT SUITE; LINE; SCIENCE; SPECTROMETRY; PARAMETERS; PRECISION; DENSITY; THOMSON

Autoři

KÉPEŠ, E.; SAEIDFIROUZEH, H.; LAITL, V.; VRÁBEL, J.; KUBELÍK, P.; POŘÍZKA, P.; FERUS, M.; KAISER, J.

Vydáno

3. 4. 2024

Nakladatel

ROYAL SOC CHEMISTRY

Místo

CAMBRIDGE

ISSN

1364-5544

Periodikum

Journal of Analytical Atomic Spectrometry

Ročník

39

Číslo

4

Stát

Spojené království Velké Británie a Severního Irska

Strany od

1160

Strany do

1174

Strany počet

15

URL

https://pubs.rsc.org/en/content/articlelanding/2024/ja/d3ja00363a

Plný text v Digitální knihovně

http://hdl.handle.net/11012/249970

BibTex

@article{BUT188826,
  author="Erik {Képeš} and Homa {Saeidfirouzeh} and Vojtěch {Laitl} and Jakub {Vrábel} and Petr {Kubelík} and Pavel {Pořízka} and Martin {Ferus} and Jozef {Kaiser}",
  title="Interpreting neural networks trained to predict plasma temperature from optical emission spectra",
  journal="Journal of Analytical Atomic Spectrometry",
  year="2024",
  volume="39",
  number="4",
  pages="1160--1174",
  doi="10.1039/d3ja00363a",
  issn="1364-5544",
  url="https://pubs.rsc.org/en/content/articlelanding/2024/ja/d3ja00363a"
}