Spectrum adapted EM algorithm for high-throughput peak fitting analysis of spectral data

Abstract

We introduce the spectrum adapted EM algorithm as a method for high-throughput analysis of spectral data. In this presentation, we explain the theoretical background of the spectrum adapted EM algorithm and introduce examples of its application to a synthetic spectral data and an experimental spectral data collected by XPS. Moreover, we introduce an advanced method of the spectrum adapted EM algorithm based on the ECM algorithm. Using this advanced method, we demonstrate recent works for the background subtraction and optimization of the number of decomposed peaks in the peak fitting of spectral data.