PATTERN

RECOGNITION

SECOND EDITION

 

SERGIOS THEODORIDIS

KONSTANTINOS KOUTROUMBAS

Pattern recognition is incredibly important in all automation, information handling and retrieval applications.  The new edition of this best-selling book, a text written by two of the field’s leading experts, covers the entire spectrum of pattern recognition applications from an engineering perspective, examining topics from image analysis to speech recognition and communications.  This thoroughly updated edition presents cutting-edge material on neural networks and highlights the latest developments in this growing field, including independent components analysis and support vector machines.  Developed through more than 10 years of teaching experience, Pattern Recognition is the most comprehensive reference available for both engineering students and practicing engineers

Coverage Includes:

Ø      Latest techniques in feature generation, including features based on Wavelets, Wavelet Packets, Fractals and a new section on Independent Component Analysis (ICA)

Ø   All new sections on Deformable Template Matching, Support Vector machines and a related appendix on Constrained Optimization

Ø       Feature selection techniques

Ø    Design of linear and non-linear classifiers, including Bayesian, Multilayer Perceptrons, Decision Trees and RBF networks

Ø  Template Matching, Context-dependent classification, including Dynamic Programming and Hidden Markov Modeling techniques

Ø     Classical approaches, as well as the most recent developments in clustering algorithms, such as fuzzy, possibilistic, morphological, genetic, and annealing techniques

Ø     Coverage of numerous, diverse applications, including Image Analysis, Character Recognition, Medical Diagnosis, Speech Recognition, and Channel Equalization

Ø     Numerous computer simulation examples, supporting the methods given in the book, available via the Web (from September 2003)

Ø    Power Point Presentation – Teaching Material available upon request