Cavanna, D., Righetti, L., Elliott, C., & Suman, M. (2018). The scientific challenges in moving from targeted to non-targeted mass spectrometric methods for food fraud analysis: A proposed validation workflow to bring about a harmonized approach. Trends in Food Science & Technology, 80, 223-241.
37. line 240 Unfortunately, ANN don’t provide any means to explain why the model is working and in FTIR spectroscopy that is a crucial issue. FTIR spectra include environmental interferences and noise in addition to chemical information. The models might have been affected by these interferences.
[Response] Thanks for your comments. There is no one-to-one correspondent relationship between these chemical indexes (polar compounds, free fatty acids, hydroperoxide) and the selected peak areas, therefore they may be not absolutely the linear relationships between the chemical indexes and the selected peaks. Many researches also used the ANN to calibrate the model with the spectra and other chemical index indicating it is available for choosing ANN for this spectroscopic application (Özbalci, et al., 2013; Wenning et al., 2010; Barmpalexis, 2018; Li, et al., 1999; Agatonovic-Kustrin, 2013).
FTIR have many merits, FTIR has the characteristics of measurable spectral band width, high spectral resolution, high signal-to-noise ratio, and fast scanning speed. The FTIR uses few optical components, no grating or prism beam splitter, which reduces the loss of light, and further increases the light signal through interference, so the radiation intensity reaching the detector is high and the signal-to-noise ratio is high. Besides, the FTIR has good reproducibility,the Fourier transform used by the FTIR processes the light signal, avoiding the error caused by the motor driving the grating to split light, so the reproducibility is better. These merits might reduce the interferences from environmental and noise.