E: edvard.govekar@fs.uni-lj.si
T: +386 1 4771-606 Primož Potočnik, asst. prof.
E: primoz.potocnik@fs.uni-lj.si
T: +386 1 4771-167
Course description |
Lectures |
Practical examples |
Seminars |
Online resources |
E‑books |
Data |
Package
Course description
- Introduction to Neural Networks
what is a neural network?, biological neural networks, human nervous system, artificial neural networks, benefits of neural networks, brief history of neural networks, applications of neural networks - Neuron Model, Network Architectures and Learning
neuron model, activation functions, network architectures, learning algorithms, learning paradigms, learning tasks, knowledge representation, neural networks vs. statistical methods - Perceptrons and Linear Filters
perceptron neuron, perceptron learning rule, adaline, LMS learning rule, adaptive filtering, XOR problem - Backpropagation
multilayer feedforward networks, backpropagation algorithm, working with backpropagation, advanced algorithms, performance of multilayer perceptrons - Dynamic Networks
historical dynamic networks, focused time-delay neural network, distributed time-delay neural network, NARX network, layer recurrent network, computational power of dynamic networks, learning algorithms, system identification, model reference adaptive control - Radial Basis Function Networks
RBFN structure, exact interpolation, radial basis functions, radial basis function networks, RBFN training, RBFN for pattern recognition, comparison with multilayer perceptron, probabilistic networks, generalized regression networks - Self-Organizing Maps
self-organization, self-organizing maps, SOM algorithm, properties of the feature map, learning vector quantization - Practical Considerations
preparing data, selection of inputs, data encoding, principal component analysis, invariances and prior knowledge, generalization, general guidelines - Advanced topics
optimal network architectures, evolution of neural networks, support vector machines, committee machines, stochastic machines, principal component networks, bayesian neural networks
Lectures
Download the PDF lectures (password protected):
| PDF download | Description |
| NN-00.pdf | 0. Organization of the Study |
| NN-01.pdf | 1. Introduction to Neural Networks |
| NN-02.pdf | 2. Neuron Model, Network Architectures and Learning |
| NN-03.pdf | 3. Perceptrons and Linear Filters |
| NN-04.pdf | 4. Backpropagation |
| NN-05.pdf | 5. Dynamic Networks |
| NN-06.pdf | 6. Radial Basis Function Networks |
| NN-07.pdf | 7. Self-Organizing Maps |
| NN-08.pdf | 8. Practical Considerations |
| NN-lectures.pdf | Complete lectures in a single file [8 MB] |
Practical examples (MATLAB)
nn02_neuron_output - Calculate the output of a simple neuron
nn02_custom_nn - Create and view custom neural networks
nn03_perceptron - Classification of linearly separable data with a perceptron
nn03_perceptron_network - Classification of a 4-class problem with a 2-neuron perceptron
nn03_adaline - ADALINE time series prediction with adaptive linear filter
nn04_mlp_xor - Classification of an XOR problem with a multilayer perceptron
nn04_mlp_4classes - Classification of a 4-class problem with a multilayer perceptron
nn04_technical_diagnostic - Industrial diagnostic of compressor connection rod defects [data2.zip]
nn05_narnet - Prediction of chaotic time series with NAR neural network
nn06_rbfn_func - Radial basis function networks for function approximation
nn06_rbfn_xor - Radial basis function networks for classification of XOR problem
nn07_som - 1D and 2D Self Organized Map
nn08_tech_diag_pca - PCA for industrial diagnostic of compressor connection rod defects [data2.zip]
| PDF download | Description |
| NN-examples.pdf | Complete Matlab examples in a single PDF file [2 MB] |
Seminars
Seminar examples
‑ Characteristic of the critical flow Venturi nozzle mass flow meter using RBFN
‑ Krmiljenje manipulatorjev na osnovi strojnega vida
‑ SONN Application
‑ Classification of 4-class spiral problem by SVM methodology
‑ Character Recognition
‑ A Tutorial on Support Vector Machine
‑ Classification of Iris data set
Online resources
- MATLAB Neural Networks Toolbox (User's Guide) -- latest version
- Artificial Neural Networks on Wikipedia.org
- Principal components analysis -- on Wikipedia.org
E-books
- An Introduction to Neural Networks, Ben Krose & Patrick van der Smagt, 1996
Krose1996.pdf (1.2 MB) - Neural Networks and Deep Learning, free online book by Michael Nielsen, 2014
Data
Package Fundamentals of Neural Networks
The open-source teaching package contains modular contents for the introduction of the fundamentals of Neural Networks. The package consists of a series of MATLAB Live Scripts with complementary PowerPoint presentations.