Theory of Principal Components Analysis
Principal Components Analysis is a widely used dimension reduction tool which has found extremely wide application. It is nonparametric and easy to implement - qualities that helped it become a mainstay of modern statistics. Yet, despite its simple definition, the standard estimator is a very complicated statistical object, and some of its aspects remain poorly understood.
This talk will present an overview of PCA and its statistical properties. Some extensions of the method will also be discussed. No prior knowledge of the subject should be needed.