Course details
Neural Networks
NEU Acad. year 2004/2005 Winter semester 6 credits
Artificial neuron, basis and activation functions. Classification of neural networks. Principles of individual neural networks (topology, learning, responses, typical applications): "Adaline, Perceptron, Madaline, BPN, adaptive feedforward multilayer networks, self-organizing neural networks, CPN, LVQ, RBF, RCE, Hopfield neural networks, BAM, SDM, Boltzmann machine, Neocognitron". Genetic algorithm, fuzzy systems, rough sets and neural networks.
Guarantor
Language of instruction
Completion
Time span
- 39 hrs lectures
- 26 hrs projects
Department
Subject specific learning outcomes and competences
Students acquire knowledge of particular types of neural networks and so they will be able to design programs using these networks to solving of various practical problems.
Learning objectives
To give the students the knowledge of fundamentals of neural network theory and the knowledge of topologies, learning, responses and possible practical applications of various types of these networks.
Prerequisite knowledge and skills
There are no prerequisites
Study literature
- Šíma,J., Neruda,R.: Teoretické otázky neuronových sítí, MATFYZPRESS, 1996, ISBN 80-85863-18-9
- Novák,M. a kol.: Umělé neuronové sítě, C.H. Beck, 1998, ISBN 80-7179-132-6
Fundamental literature
- Mehrotra,K., Mohan,C.K., Ranka S: Artificial Neural Networks, The MIT Press, 1997, ISBN 0-262-13328-8
- Hassoun, M.H.: Artificial Neural Networks, The MIT Press, 1995, ISBN 0-262-08239-X
- Haykin,S.: Neural Networks, Macmillan College Publishing Company, Inc., 1994, ISBN 0-02-352761-7
Syllabus of lectures
- Introduction, artificial neuron, classification of neural networks
- Perceptron, Adaline, Madaline
- Bacpropagation (BP) Neural Network
- Constructive neural networks
- RBF and RCE neural networks
- Topologic organized neural network, CPN, LVQ
- ART and SDM neural networks
- Neural networks as associative memories, Hopfield network, BAM
- Optimization problems solving using neural networks, Stochastic neural networks, Boltzmann machine
- Neocognitron neural network
- Genetic algorithm and neural networks
- Fuzzy systems and neural networks
- Rough sets and neural networks
Progress assessment
Study evaluation is based on marks obtained for specified items. Minimimum number of marks to pass is 50.
Controlled instruction
There are no checked study.