Prediction of milk composition using multivariate chemometric modelling of spectral data and data fusion approaches

The first chapter of the thesis is a brief introduction to the whole study. In the second chapter, a literature survey has been performed on available sources to explore the spectroscopic analysis of liquid milk and all the potential data analysis approaches applied in this way. The third, chapter focuses on the classification of different types of milk using NIR, FT-IR, and Raman spectroscopy and the application of PCA, PLS-DA, SO-PLS-LDA, and machine learning methods. In Chapter 4, the profile of fatty acids in liquid milk has been explored with NIR and FT-IR spectroscopy and PLSR. Data fusion as a newly emerged data analysis approach has also been investigated in chapters 2, 3, and, 4 to employ the potential of this strategy in improving the results.