Autonomous Agents and Multi-Agent Systems (2023/2024)

Course code
cod wi: DT000457
Name of lecturer
Alessandro Farinelli
Alessandro Farinelli
Number of ECTS credits allocated
Academic sector
Language of instruction
Anno accademico 2023/2024 Dottorato di Ricerca dal Oct 1, 2023 al Sep 30, 2024.

Lesson timetable

Go to lesson schedule

Learning outcomes

The course illustrates the main issues related to the development of intelligent agents that can perceive, plan, act and interact with other agents and humans. The goal is to provide students with tools to design, apply and evaluate algorithms that allow intelligent agents to interact with the surrounding environment by performing complex tasks with a high level of autonomy. At the end of the course, students will have to demonstrate that they understand the fundamental concepts related to: i) Cooperative Artifical Intelligence and in particular optimization in multi-agent contexts; ii) Machine Learning with emphasis on Reinforcement Learning; iii) Artificial Intelligence techniques for robotic systems. Students will have to demonstrate knowledge and be able to use the main tools for the development of autonomous agents and multi-agent systems. Students will also have to know the open challenges and limitations of state-of-the-art techniques for the area related to intelligent agents and multi-agent systems and have the ability to continue their studies independently by developing innovative approaches aimed at improve the state of the art.


i) Introduction to the broad areas of Autonomous Agents and Multi-Agent Systems and Cooperative Artificial Intelligence;
ii) algorithms and techniques to perform optimization in the context of Multi-Agent Systems (with a specific focus on Distributed Constraint Optimization approaches);
ii) techniques and methodologies to learn how to operate in uncertain and dynamic environment, with a specific focus on reinforcement learning, multi-agent reinforcement learning and safe reinforcement learning;
iii) Artificial Intelligence techniques for mobile robots and multi-robot systems (e.g., multi-robot coordination, deep reinforcement learning for robotic systems).

Assessment methods and criteria

The exam can be taken by choosing between two options: i) a project that includes an experimental part focused on the techniques studied during the course; ii) an oral presentation based on state-of-the-art articles and papers that explores some of the topics studied during the course.