Simultaneous quantification of Au and Ag composition from Au–Ag bi-metallic LIBS spectra combined with shallow neural network model for multi-output regression

Trustworthiness of Laser-Induced Breakdown Spectroscopy Predictions via Simulation-based Synthetic Data Augmentation and Multitask Learning

“…At the same time, the production of purely synthetic data, based on local thermodynamical equilibrium, has been explored in LIBS applications [10]. The idea of enriching existing data by means of different representations of the inputs, such as time resolved spectra, was experimented with success [11]. More recently, standard data augmentation techniques were used for the classification of LIBS mapping experiments [4].…”

“…Though explored at length in ML, MT learning [13] has seen major developments and a wide range of applications recently. Some examples of multi-output algorithms for LIBS analyses were recently explored, based on NNs [11] and on Partial Least Squares (PLS) with two outputs, i.e. PLS2 [14].…”

Atomic spectrometry update: review of advances in the analysis of metals, chemicals and materials

“…An example, by Narlagiri and Soma used LIBS followed by data input to PCA to reduce the number of dimensions before the output of this was fed into a new regression for quantification. 55 The new regression was a multi-output regression with shallow neural networks that uses two nodes at the output layer. These nodes calculate the composition of one element each.…”

Deep Learning Models for Data-Driven Laser Induced Breakdown Spectroscopy (LIBS) Analysis of Interstitial Oxygen Impurities in Czochralski-Si Crystals

“…39 Other efforts using neural network models for various spectroscopy data analyses include the use of convolutional neural networks (CNN) for predicting potassium concentration in soil with significantly improved accuracy, 40 as well as soil sensing techniques using 1D CNN on visible-near-IR spectroscopy, 41 and quantitative LIBS analyses of elemental compositions of bimetallic targets using shallow neural network models for multi-output regression. 42 In this study, we follow-up our previous investigation 23 to bridge the void between LIBS experimental data and data analytics using one-dimensional (1D) CNN to predict the concentration of interstitial oxygen dissolved in non-doped commercial grade Si wafers based on LIBS spectral data. We compare our results with the ones from traditional LIBS calibration techniques using O(I) (777.19 nm) peak, as well as from least absolute shrinkage and selection operator (LASSO) method.…”

“…39 Other efforts using neural network models for various spectroscopy data analyses include the use of convolutional neural networks (CNN) for predicting potassium concentration in soil with significantly improved accuracy, 40 as well as soil sensing techniques using 1D CNN on visible–near-IR spectroscopy, 41 and quantitative LIBS analyses of elemental compositions of bimetallic targets using shallow neural network models for multi-output regression. 42…”