ABSTRACT
This chapter presents some novel approaches that can be applied in food analysis, such as data fusion and multivariate curve resolution (MCR) and discusses preprocessing and unsupervised classification methods. Chemometrics is the application of multivariate statistics in chemistry and is defined as the chemical discipline that uses mathematical and statistical methods to design or select optimal measurement procedures and experiments and provide maximum chemical information by analyzing chemical data. Hierarchical cluster analysis (HCA) is an unsupervised method often employed in exploratory data analysis and pattern recognition. The Euclidean distance is the most used measure for HCA. One of the most employed methods to extract relevant information in experimental data and pattern recognition is Principal component analysis. In the milk dataset, the complete-linkage was applied, which is one of the simplest linkage criteria. In multivariate calibration, several variables per sample are available, such as in sensor arrays, spectra, voltammograms, and chromatograms.
