Introduction to Statistical Learning and Pattern Classification

Many of the chapters here concerned with experimental work (both for neurophysiological recordings and functional imaging measurements) have taken advantage of powerful machine learning techniques developed in the last several decades. Here we describe the underlying mathematics, discuss issues that are relevant to those who study the brain, and summarize current applications of these techniques. Chapters 19 and 20 extend these concepts and are concerned with the application of such techniques to neuronal and fMRI data, respectively. The material presented here is only an introduction to the topic, and the reader desiring more thorough coverage should consult more advanced textbooks (e.g., Bishop, 1995; Cristianini and ShaweTaylor, 2000; Duda and Hart, 1973; Gabbiani and Cox 2010; Hertz et al 1992; Poggio and Smale 2003; Poggio et al 2004; Vapnik, 1995). Press et al. (1996) is also an excellent reference for computational algorithms to efficiently implement these techniques.