Optimization for artificial intelligence

Learning Goals

The educational objectives of the course are:
– [Knowledge and understanding] To enable students to know and understand optimization algorithms based tu metaheuristics, particularly evolutionary algorithms and based on swam intelligence methods.
– [Applying knowledge and understanding] Make students able to understand strengths and weaknesses of the approaches of the approaches studied and which techniques to apply to specific optimization problems. Also make them able to adapt and develop variants of existing algorithms.
– [Making judgments] Enable students to understand which problems might benefit from a metaheuristic approach and then choose which techniques to adopt.
– [Communication skills] To enable students to explain to both a broad audience and a specialized audience the operating principles of metaheuristics, clearly explaining their areas of applicability and the advantages and disadvantages of their adoption.
– [Learning skills] It is intended that students be able to independently investigate and update themselves on new evolutionary and swarm intelligence techniques and their applications by making use of different sources of information (books, scientific articles, seminars, etc.).

Program in pills

The course will cover how to perform optimization through a set of bio-inspired techniques. In particular, it will cover evolutionary techniques (genetic algorithms, genetic programming) and swarm intelligence (particle swarm optimization) also combined with other machine learning methods (e.g., neuroevolution). It will show how to apply these techniques to problems where there are multiple objectives.

Area

Formazione matematico-statistica

Curriculum Foundations
TAF Type

B

Curriculum Industry
TAF Type

Curriculum Health
TAF Type

Curriculum Economy
TAF Type

SSD

MAT/09

ECTS

6

Semester

1

Lecturers

Luca Manzoni