ABSTRACT
This chapter is an introduction to the field of robust multivariate techniques for dimension reduction. We focus on principal components analysis, and its robust counterparts.
In the first section of this chapter we review classical PCA. PCA is a widely used technique for descriptive multivariate statistics and dimensionality reduction. This material will only be covered shortly, in order to keep the book self contained. The reader can refer to Jolliffe (2005) for an extensive account on classical dimension reduction.
