Functional Data Analysis for Spectroscopy Data

We propose a functional data analysis approach for the study of spectroscopy data. The applicative problem concerns the characterization of the spectral response of silicate glasses in order to infer the chemical composition of the materials from spectral data, which can be remotely collected. The analysis performed aims at characterizing the phase variability and the amplitude variability of the data in order to extract meaningful information. The technique used is k-mean alignment.