Chapter 8 - Principal Components Analysis

Abstract

This chapter presents principal components analysis (PCA), which is a multivariate technique used to describe the relationship between several response variables and to explain the total variation in the data. PCA uses a few equations constructed from the original variables, which are called components. PCA is very useful when the variables under study are highly correlated (positively or negatively) or when the number of independent variables is large. Calculations of PCA and a step-by-step procedure without complicated mathematical proofing is given to enable the reader to understand the concept of PCA and how to apply it to food science. Two examples are given from food science and analyzed by PCA with R commands; the first example addresses the physicochemical properties of bananas, and the second example considers the antioxidants’ activity and antioxidative compounds.