Human communication is multimodal per se, that is human effectively communicate meaning, feelings and emotions through multiple sensory channels such as speech, movement, and facial expressions. A lot of effort has been devoted to conceive intelligent machines able to capture, represent, and automatically analyze the multimodal behavior of their users in order to engage them in a multimodal dialog aimed at establishing a natural interaction. This course provides students with foundational conceptual knowledge, methodologies, and tools for designing, implementing, and evaluating such kind of intelligent machines.

This course will present some basic learning algorithms (linear regression, logistic regression, SVM, Bayesian Networks, etc.). Students will have the chance to implement and try some of these algorithms on simple examples. Very wide application fields in robotics and social robotics will be presented.

Introductory level class on Machine Learning on generic data: Supervised Learning, Unsupervised Learning, Kernels, Neural Networks, Basics of Deep learning and Graph neural network.

This course introduces students to the advanced principles of Deep Learning, including mathematical foundations, architecture design and practical applications. This course is particularly relevant given the current state of the job market, where Deep Learning skills are in high demand in many sectors, including technology, finance, healthcare and entertainment. The class include the following topics: GAN, Attention and Trasnformers, Graph Neural Networks, Neural architect search, and various interdisciplinary applications: Large Pretrained Models, Vision & Language, Biological Applications, Time Series.

This course selects a number of advanced topics to explore in machine learning and autonomous agents, in particular: Probabilistic graphicsal models (Bayesian networks), Multi-output and structured-output prediction problems, deep-learning architectures, Methods of search and optimization, Sequential prediction and decision making, Reinforcement learning.