1. SylabUZ
2. Faculty of Engineering and Technical Sciences
3. winter term 2018/2019
4. WIEiA - oferta ERASMUS / Informatics - Second-cycle Erasmus programme
5. Neural and neuro-fuzzy networks

Generate PDF for this page

Neural and neuro-fuzzy networks - course description

Aim of the course

• Familiarize students with different architectures of artificial neural networks and neuro-fuzzy systems.
• Familiarize students with the learning methods of neural networks and neuro-fuzzy systems.
• Development of skills to use neural networks and neuro-fuzzy networks for modeling and pattern recognition.

 

Prerequisites

Save changes

Scope

Feedforward neural networks. Fundamentals of multilayer neural networks. Backpropagation algorithm for neural network learning. Issues and limitations of gradient descent learning algorithms. Adaptive learning rate. Momentum. Review of advanced learning algorithms. Sample applications of neural networks.

Recurrent neural networks. Dynamic-feedback neural networks. Locally recurrent globally feedforward networks. Learning algorithms for feedback neural networks. Hopfield neural network. Learning algorithms for Hopfield neural network. 

Self-organizing neural networks. Kohonen self-organizing feature maps. Competitive learning.  Algorithm of neural gas. Sample applications of the Kohonen network.

Deep learning. Convolutional neural network. Restricted Boltzman Machine. Deep Belief Networks. Autoencoders. Fast deep learning with GPU computations.

Neuro-fuzzy systems. Fuzzy sets and fuzzy logic. Operations on fuzzy sets. Fuzzy inference. Fuzzy rules. Mamdani and Takagi-Sugeno neuro-fuzzy systems. Gradient descent based learning algorithm for neuro-fuzzy systems.  

Teaching methods

Lecture, teaching laboratory classes.

 

Learning outcomes and methods of theirs verification

Outcome description	Outcome symbols	Methods of verification	The class form

Assignment conditions

Lecture - the passing criterion is a sufficient mark from the final test.
Laboratory - the passing criterion are positive marks for laboratory exercises and tests.
Final mark components = lecture: 50% + teaching laboratory: 50%

Recommended reading

1. Haykin S.: Neural Networks: A Comprehensive Foundation (2nd Edition), Prentice Hall, 1998..
2. Goodfellow I., Bengio Y., Courville A.: Deep Learning, MIT Press, 2016.
3. Bishop C.M., Hinton G.: Neural Networks for Pattern Recognition, Clarendon Press, Oxford, 1995.
4. Zimmermann H-J.: Fuzzy Set Theory and Its Applications, Springer, 2006.
5. Rutkowska D.:Neuro-Fuzzy Architectures and Hybrid Learning, Springer, 2001.
6. Pal S.K., Mitra S.: Neuro-Fuzzy Pattern Recognition: Methods in Soft Computing, Wiley, 1999.

Further reading

1. Murphy K.P.: Machine Learning. A Probabilistic Perspective, MIT Press, 2013.
2. Hastie T., Tibshirani R., Friedman J: The Elements of Statistical Learning, Springer, 2001.
3. Theodoridis S.: Machine Learning. A Bayesian and Optimization Perspective. Academic Press, 2015. 
4. Russell S., Norvig P.: Artificial Intelligence: A Modern Approach, Prentice Hall, 2009.

Notes

Modified by dr hab. inż. Marek Kowal, prof. UZ (last modification: 27-03-2018 18:09)