13M051MSC - Soft-computing Methods
|Electrical Engineering and Computing
|Type of study
|master academic studies
|Lecturer (for classes)
|Lecturer/Associate (for practice)
|Lecturer/Associate (for OTC)
|Objective of the course is for the students to become able to design neural networks, fuzzy systems and genetic algorithms with the final aim to implement them in process control, signal processing and pattern recognition problems.
|Students will obtain the skills to choose the adequate soft-computing technique, define necessary a priori knowledge, generate or acquire corresponding algorithm training set, make a fine tuning of controlling parameters, design an algorithm and evaluate it, depending on the problem at hand.
|URL to the subject page
|URL to lectures
|Contents of lectures
|Structure of genetic algorithms; Selection, crossing-over, mutation; Optimization of criteria with constrains based on genetic algorithms; Travelling salesman problem; Fuzzy sets; Fuzzy operation; Fuzzy algorithms; Design of fuzzy expert systems; Introduction to neural networks; Types of neural networks; Backward error propagation; Application of neural networks; Associative memory.
|Contents of exercises
|Design of particular fuzzy expert system, design of fuzzy controllers and application of fuzzy algorithm in pattern recognition; Optimization of particular problem using genetic algorithm; Design of process controller or classification system using on neural networks.
|Number of hours per week during the semester/trimester/year
|Study and Research
|Methods of teaching
|Lectures and practical lab work
|Knowledge score (maximum points 100)
|Activites during lectures