Evaluating different deep learning models for efficient extraction of Raman signals from CARS spectra

Abstract

The non-resonant background (NRB) contribution to the Coherent anti-Stokes Raman scattering (CARS) signal distorts the spectral line shapes and thus degrades the chemical information. Hence, finding an effective approach for...

Log-Gaussian gamma processes for training Bayesian neural networks in Raman and CARS spectroscopies

“…A comparison study similar to 35 would be an interesting avenue of future research. However, a direct comparison between non-Bayesian and Bayesian neural networks is not straight-forward due to the missing uncertainty quantification of the non-Bayesian neural network estimates.…”

“…24–28 Similarly, they have been applied to extract the underlying Raman spectra from CARS measurement. 29–35 Despite their efficacy, non-Bayesian neural networks lack a critical feature: the ability to quantify uncertainty in Raman spectrum estimation. Bayesian inference, on the other hand, provides an avenue to solve this problem.…”

A deep spectral prediction network to quantitatively determine heavy metal elements in soil by X-ray fluorescence

“…Feature engineering includes principal component analysis (PCA) 14 and competitive adaptive reweighted sampling (CARS). 15 Regression algorithms are support vector regression (SVR) 16 and random forest regression (RFR). 17 Statistical methods usually refer to partial least squares regression (PLSR).…”