PORTALE DELLA DIDATTICA

PORTALE DELLA DIDATTICA

PORTALE DELLA DIDATTICA

Elenco notifiche



Advanced techniques for optimization

01QFFRV

A.A. 2020/21

Course Language

Inglese

Degree programme(s)

Doctorate Research in Ingegneria Elettrica, Elettronica E Delle Comunicazioni - Torino

Course structure
Teaching Hours
Lezioni 15
Esercitazioni in aula 5
Lecturers
Teacher Status SSD h.Les h.Ex h.Lab h.Tut Years teaching
Pirinoli Paola Professore Ordinario IINF-02/A 15 5 0 0 7
Co-lectures
Espandi

Context
SSD CFU Activities Area context
*** N/A ***    
PERIOD: January-February The course aims to introduce several global optimization techniques and to discuss their application to different engineering problems. Even if the proposed approaches are general purpose techniques and could in principle be applied to the optimization of any kind of problems, after the study of several well-known or more innovative methods their behavior will be analyzed when applied to different problems, to enlighten their strengths and their drawbacks. The course will give both the competences on different optimization techniques, and on their use for the optimization of complex problems, as for instance the design of microwave components. The course outcome will be not only the knowledge of the principal and more innovative global optimization methods, but also the capability of choosing the proper approach for a given problem and to mathematically model this last in the most suitable way. The course is organized in theoretical classes, during which the considered algorithms are presented, and labs, during which the students are asked to apply some of the analyzed techniques to the optimization of several benchmark functions, with the aim to become familiar with them, to study the effects of their characteristic parameters on their behavior and to compare their performance on different classes of problems. The acquired competences will be tested through the development of a project, that could be carried on in small groups or individually, focused on the application of one or more of the considered optimization techniques to a problem of their own interest. The results are then presented to the other course attendees during a dedicated class.
The course is devoted to all the Engineering PhD students, and therefore any particular prerequirement is asked, except the Matlab programming capability.
Introduction to optimization: global vs. local, deterministic vs. non-deterministic techniques Global optimization methods: evolutionary approaches Genetic Algorithms Particle Swarm Optimization Innovative techniques: BBO, SNO, ... Hybrid methods Multi-objective approaches Introduction to neural networks Estimation distribution algorithms: Bayesian Optimization Algorithm, Compact Genetic Algorithm
Mixed mode
Team project work development
P.D.1-1 - January