Pattern Recognition

Methods and pattern recognition systems, Limitation in accuracy of recognition reliability, Guided learning and self-learning, Distance Functions. Linear and non-linear decision functions, Perceptron Algorithm, Bayes Classifiers, Nearest neighbor classifiers, Parametric and non-parametric estimation of probability density models: Maximizing entropy, Parzen estimator, orthonormal functions, Robbins Monro and Kiefer Wolfowitz methods, LMS, Least squares Methods., Multilayer artificial neural networks, Recursive artificial neural networks, Error correction training, Hebbian and competitive training, Multilayer perceptron, Error Back Propagation, Radial basis function networks, Hopfield machine, supervised and unsupervised learning, Hierarchical data clustering, Fuzzy logic, Genetic algorithms and evolutionary computation principles.

Objectives

The student will have the ability to understand the meaning of the term "Pattern recognition" as a concept of categorization of objects into classes. He/She will also be familiar with Bayes classifiers and classification procedures (normal distribution and classification) and will have faced reduction of feature vectors, control case and dispersion, feature extraction in programming environment Matlab.

Prerequisites

Non required

Syllabus

Methods and pattern recognition systems, Limitation in accuracy of recognition reliability, Guided learning and self-learning, Distance Functions. Linear and non-linear decision functions, Perceptron Algorithm, Bayes Classifiers, Nearest neighbor classifiers, Parametric and non-parametric estimation of probability density models: Maximizing entropy, Parzen estimator, orthonormal functions, Robbins Monro and Kiefer Wolfowitz methods, LMS, Least squares Methods., Multilayer artificial neural networks, Recursive artificial neural networks, Error correction training, Hebbian and competitive training, Multilayer perceptron, Error Back Propagation, Radial basis function networks, Hopfield machine, supervised and unsupervised learning, Hierarchical data clustering, Fuzzy logic, Genetic algorithms and evolutionary computation principles.