Courses

An Introduction to Reinforcement Learning: from Fundamentals to AI Agentic Systems

This course provides an introduction to reinforcement learning (RL), guiding participants from core concepts to the emerging field of AI agentic systems. We will introduce the fundamental principles that enable intelligent agents to learn through interaction with their environment, and we will explore the role of Markov Decision Processes (MDPs) in modelling sequential decision-making problems. Building on these foundations, we will discuss recent advances in deep reinforcement learning and the development of autonomous, goal-directed AI agents. Through conceptual explanations and illustrative examples, participants will gain insight into how RL algorithms enable systems to adapt, optimise decisions, and operate in complex environments. By the end of the course, participants will have a clear understanding of the basic theory behind reinforcement learning and its growing importance in the design of intelligent and agentic AI systems, providing a foundation for further study or practical experimentation in this rapidly evolving field.

Instructor: Mirco Musolesi is Professor of Computer Science at the Department of Computer Science at University College London, where he leads the Machine Intelligence Lab. He is also Professor of Computer Science at the Department of Computer Science and Engineering at the University of Bologna, Italy. Previously, he held research and teaching positions at Dartmouth, Cambridge, St Andrews, and Birmingham. He has broad research interests spanning several traditional and emerging areas of Computer Science and beyond. More specifically, current research areas of his lab include foundations of Machine Learning/Artificial Intelligence (with a focus on decision making in single-agent and multi-agent environments and generative artificial intelligence), computational modeling of user/human/machine behavior, AI&cybersecurity and AI&society. More information about his research profile can be found at https://www.mircomusolesi.org/.

The goal of this course is to introduce Artificial Social Intelligence, commonly referred to as Social AI, a field focused on the automatic modelling, analysis, and synthesis of verbal and non-verbal behaviour in both human–human and human–machine interactions. After presenting the field and its key scientific and technological objectives, the course will cover the main methodological steps underlying Social AI, from the collection of behavioural data to the design of appropriate AI approaches and the psychologically informed evaluation of results. Special attention will be given to the interdisciplinary nature of Social AI. The course will demonstrate how social and psychological questions can be formulated in technological terms and, conversely, how addressing technological challenges related to human behaviour can generate valuable psychological insights.

Instructor: Alessandro Vinciarelli is a Full Professor at the University of Glasgow, where he is affiliated with the School of Computing Science and the Institute of Neuroscience and Psychology. His main research interest is Social AI, a field focused on the modelling, analysis, and synthesis of verbal and non-verbal behaviour in human–human and human–machine interactions. He has published more than 200 papers in international journals and conferences and has served as Principal Investigator on more than 15 research projects. These include an EU-funded European Network of Excellence, SSPNet (2009–2014, €6.2M), and the UKRI Centre for Doctoral Training in Socially Intelligent Artificial Agents (2019–2027, €6M, http://socialcdt.org). Alessandro has also co-organized more than 30 international events, serving as General Chair (e.g., IEEE International Conference on Social Computing 2012, ACM International Conference on Multimodal Interaction 2017, International Conference on Digital Mental Health and Wellbeing 2026), Program Chair (ACM International Conference on Multimodal Interaction 2023, Affective Computing and Intelligent Interaction 2024), and in other organizational roles. In addition, Alessandro is a co-founder of Klewel, a knowledge management company recognised by the IEEE as an exemplary impact story, and he serves as a scientific advisor to Substrata, a leading company in Social Signal Processing.

Harnessing Parallel Computing Resources via Structured Programming Models

The lessons discuss the recent, impressive improvements in the parallel computing scenario, both from the hardware perspective and from the software perspective. State-of-the-art as well as structured parallel programming models will be discussed, along with the associated possibilites, problems and perspectives. If the case, exploitation of available AI support will be investigated. Lesson will be therefore divided into three parts:

hardware adavances (general purpose multicores and accelerators, AI driven hardware accelerators)
state-of-the-art parallel programming models and frameworks (principles, hands-on, perspectives)
structured parallel programming models (principles, hands-on, percolation into state-of-the-art-toos, perspectives)

Instructor: Prof. Marco Danelutto is a full professor at the Univ. of Pisa, Dept. of Computer Science. His main research interests are in the field of structured parallel programming models for parallel and distributed architectures and include design of parallel programming frameworks, tools to support parallel program development, autonomic management of non functional features, software components, parallel design patterns and algorithmic skeletons. Danelutto actively participates in the design and development of FastFlow, a structured, highly efficient, parallel programming framework targeting heterogeneous multi/many core architectures. Recently, his research interests included programming abstractions supporting the targeting of hardware accelerators. He participated and participates in different national and international research projects and he is the author/co-author of more than 200 papers in international refereed journals and conferences.