New opportunities in solution of inverse problems in laser spectroscopy due to application of artificial neural networks

Abstract

Inverse problems in laser spectroscopy are reported to be successfully solved by application of artificial neural networks, a powerful data processing technique that performs better than traditional variation algorithms of solving inverse problems.Keywords: laser spectroscopy, inverse problems, artificial neural networks 157 Downloaded From: http://proceedings.spiedigitallibrary.org/ on 06/22/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx Proc. SPIE Vol. 4749 158 Downloaded From: http://proceedings.spiedigitallibrary.org/ on 06/22/2016 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

Raman Spectroscopy of Water–Ethanol Solutions: The Estimation of Hydrogen Bonding Energy and the Appearance of Clathrate-like Structures in Solutions

“…Figures – and Figure S1 show the obtained polarized Raman spectra of water, ethanol, and water–ethanol solutions at different temperatures. The behavior of the Raman stretching band of water with a temperature is similar to the data obtained by many authors. − …”

Evidence of carbon nanoparticle–solvent molecule interactions in Raman and fluorescence spectra

“…For this estimation, the analysis of position and shape of the water, Raman valence band in water suspensions of NDs was used. It comes as no surprise because as it was shown many times the Raman valence band of vibrations of OH‐groups is very sensitive to the presence of different admixtures in water even in small amounts . NDs in water suspensions induce changes in the structure of surrounding water .…”

Monitoring of nanodiamonds in human urine using artificial neural networks

“…When ANN are used to solve inverse problems in optical spectroscopy, three approaches can be used: “model‐based,” “quasi‐model,” “experiment‐based” . The “model‐based” approach requires an analytical model of solution of the direct problem, to compute the data set for ANN training.…”

“…ANN is one of the most powerful algorithms of data analysis providing efficient solution of pattern recognition problems and of multi‐parameter inverse problems in many areas, including optical spectroscopy . Adaptive methods of data analysis are used today in almost all fields of bioinformatics , as here there is often lack of adequate physical or mathematical models describing the studied objects.…”