A root-mean-square-error analysis of two-peak Gaussian and Lorentzian fittings of thin-film carbon Raman spectral data

Abstract

We perform two-Gaussian and two-Lorentzian peak fits for Raman spectral data corresponding to samples of thin-film carbon found in the scientific literature. We find that the “goodness-of-fit,” as determined through an evaluation of the root-mean-square-error, is best for the two-Gaussian peak case over most of the thin-film carbon genome, except for the graphitic carbon and ta-C:H regions. We speculate that this is related to the lower levels of disorder present within the graphitic carbon and ta-C:H regions.

Evaluation comparison of wave amount measurement results in brass-plated tire steel cord using RMSE and cosine similarity

“…Calculations that will be predicted usually produce a material consideration in a large evaluation. RMSE has been adapted in many recommendation systems [28], such as evaluating the model fit of diffusion [29], thin-film carbon Raman spectral data [30], mitigating wet bias of soil moisture ocean salinity [31], surface accuracy evaluation of large deployable mesh reflectors [32]. If the RMSE value is low, then the value of a prediction that will be generated in a model or design will be close to the value in the equation.…”