An automated approach for analysis of Fourier Transform Infrared (FTIR) spectra of edible oils

Siong Fong Sim, Woei Ting
Talanta, Volume 88, January 2012

Abstract

This paper reports a computational approach for analysis of FTIR spectra where peaks are detected, assigned and matched across samples to produce a peak table with rows corresponding to samples and columns to variables. The algorithm is applied on a dataset of 103 spectra of a broad range of edible oils for exploratory analysis and variable selection using Self Organising Maps (SOMs) and t-statistics, respectively. Analysis on the resultant peak table allows the underlying patterns and the discriminatory variables to be revealed. The algorithm is user-friendly; it involves a minimal number of tunable parameters and would be useful for analysis of a large and complicated FTIR dataset.

Go to Journal

Check Also

Predicting Soil Carbon Gains from Biochar Application Using Interpretable Machine Learning

Significance  Reference Ray, A., X. Li, Y. Chen, X. Yang, W. Zhang, and D. Hui. …