The aim of this book is to serve pedagogic goals as a complement of the book Pattern Recognition, 4th Edition, by S. Theodoridis and K. Koutroumbas (Academic Press, 2009). It is the offspring of our experience in teaching pattern recognition for a number of years to different audiences such as students with good enough mathematical background, students who are more practice-oriented, professional engineers, and computer scientists attending short training courses. The book unravels along two directions.

The first is to develop a set ofMATLAB-based examples so that students will be able to experiment with methods and algorithms met in the various stages of designing a pattern recognition system—that is, classifier design, feature selection and generation, and system evaluation. To this end, we have made an effort to “design” examples that will help the reader grasp the basics behind each method as well as the respective cons and pros. In pattern recognition, there are nomagic recipes that dictatewhichmethod or technique to use for a specific problem. Very often, old good and simple (in concept) techniques can compete, from an efficiency point of view, with more modern and elaborate techniques. To this end, that is, selecting the most appropriate technique, it is unfortunate that these days more and more people follow the so-called black-box approach: try different techniques, using a related S/W package to play with a set of parameters, even if the real meaning of these parameters is not really understood.