Improving laser-induced breakdown spectroscopy regression models<i>via</i>transfer learning
Erik Képeš, Jakub Vrábel, Pavel Pořízka, Jozef Kaiser
Abstract
Calibration datasets can be extended by using artificial neural networks to transform data collected on different LIBS systems. Hence, regression performance can be improved.
Topics & Concepts
Laser-induced breakdown spectroscopyCalibrationArtificial neural networkRegressionSpectroscopyTransfer of learningRegression analysisBiological systemLinear regressionMachine learningMaterials scienceAnalytical Chemistry (journal)Computer scienceArtificial intelligenceChemistryMathematicsStatisticsPhysicsChromatographyBiologyQuantum mechanicsLaser-induced spectroscopy and plasmaAnalytical chemistry methods developmentMercury impact and mitigation studies