At the end of the course, the student will have acquired: - Knowledge of the main characteristics of Artificial Intelligence: historical overview and modern definition; Roles of Machine and Deep Learning in the broader field of AI; - Knowledge of the main Machine Learning paradigms: Supervised Learning, Unsupervised Learning, and Reinforcement Learning; - Knowledge of the main characteristics of Statistical Machine Learning: theory and algorithms; - Knowledge of the main characteristics of Artificial Neural Networks and modern Deep Learning techniques; - Knowledge of the main characteristics of modern Transformers architectures; - First-hand experience on how to read, understand, present, and put into practice the methods presented in advanced research papers in the field; - Hands-on experience in designing, implementing, and evaluating a full machine learning pipeline and on state-of-the-art deep learning methods, using state-of-the-art open-source tools and public datasets.

- Basic probability and statistics - Linear algebra - Python programming skills

- Artificial Intelligence: historical definition, brief overview, modern definition and current role of machine and deep learning. - Basic probability and statistics - Naïve Bayes Classifier - Generative and discriminative methods - Basics of learning theory - Unsupervised Learning; Basic algorithms: K-Means, Gaussian Mixture Models - k-Nearest Neighbors - The Perceptron - Artificial Neural Networks - Optimization and regularization techniques - Support Vector Machines - Kernel methods - Introduction to Deep Learning - Convolutional Neural Networks - Sequence models: Recurrent Neural Networks, LSTMs, Attention Mechanisms, and Transformers - Visualizations - Self-supervised Learning and Generative Models - Distributed and Federated Learning - Semantic Segmentation - 3D Learning - Reinforcement learning

The course includes class-taught lectures, practice laboratories, group projects, and reading groups. Lectures cover basic machine learning principles, rooted in probability theory and statistical learning theory, key algorithms, modern deep learning architectures, including Convolutional Neural Networks and Transformers, and advanced techniques. Individual laboratory sessions cover the lecture topics of the first part of the course with a hands-on approach, providing students with the necessary practice for handling complex data, implementing and training deep learning pipelines, and analyzing the results. Labs are not graded, but the acquired skills are an essential prerequisite for successfully developing the group project. In the group project activities, students are divided into groups and work on specific projects selected from a set of advanced topics. Guided by teaching assistants, the groups explore and experiment with machine and deep learning methods on challenging problems. Finally, each group is also assigned a research paper to read, understand, and present in open reading group sessions. This activity allows students to acquire the necessary skills to distill relevant knowledge from the fast-paced machine and deep learning research literature, and learn how the field is evolving.

Copies of the slides used during the lectures will be made available. All teaching material is downloadable from the teaching Portal. Reference books: - “Machine Learning: a probabilistic perspective”; K. P. Murphy - “Deep Learning”; Ian Goodfellow, Yoshua Bengio, Aaron Courville https://www.deeplearningbook.org/ - “Understanding Machine Learning: From Theory to Algorithms”; Shai Shalev-Shwartz and Shai Ben-David - “Foundations of Computer Vision”; Antonio Torralba, Phillip Isola, William Freeman https://visionbook.mit.edu/ - “Artificial Intelligence: A Modern Approach”; Stuart J. Russel and Peter Norvig

Lecture slides; Text book; Lab exercises; Video lectures (current year); Student collaboration tools;

Exam: Written test; Compulsory oral exam; Individual project; Group project;

The exam is composed of the following three parts: 1) Group paper presentation: Groups are assigned a key research paper to read and present in class during reading group sessions. Following the presentation, students answer oral questions on the paper. No notes are allowed while presenting. The evaluation is individual. - Minimum individual points: 6 - Maximum individual points: 10 (+) Passing the project presentation evaluation is a necessary prerequisite for taking the written exam. 2) Group project: Each group is assigned a project based on their preference from a list of advanced topics. Groups carry out all the mandatory steps of the project and submit a report presenting and discussing the addressed problem, related works, employed methodologies, and results. The report is then graded. The grade is collective. - Minimum group points: 3.5 - Maximum group points: 6 Passing the project report evaluation is a necessary prerequisite for taking the written exam. 3) Written exam: The written exam is individual, and consists of multiple questions evenly covering the basic and advanced topics presented during the lectures. Students have around 2.5 hours to answer a set of open questions (4 to 7, depending on the topic and complexity). - Minimum individual points: 9 - Maximum individual points: 15 The final grade is composed of the three points above, as follows: - Written (15) + Project (6) + Paper Presentation (10+) - Maximum grade: 31+ - Laude: Grade > 30 and unanimity of all examiners After the paper presentation and project report submission, students will receive a pass/not pass notification for each part. Passing both the paper presentation and the group project is a necessary requirement for taking the written exam. The final grade will be communicated after the written exam evaluation, taking into account potential bonuses and roundings. The laude is assigned to students with a total score higher than 30 points and with the unanimous agreement of the exam commission. All students are expected to adhere to the University's ethical code. If irregularities are found, the exam will be automatically withdrawn, and proper actions will be taken. - The use of generative AI tools for the preparation of slides and reports is strongly discouraged. - Plagiarism (report, slides, and/or code) will not be tolerated. - During the exam, books, notes, laptops, tablets, mobile phones, etc., are strictly forbidden. Only pen and paper. Any irregularities in the written exam will not be tolerated. All students breaking those rules will be reported to the disciplinary committee. These measures are ultimately aimed at ensuring an effective learning environment.

In addition to the message sent by the online system, students with disabilities or Specific Learning Disorders (SLD) are invited to directly inform the professor in charge of the course about the special arrangements for the exam that have been agreed with the Special Needs Unit. The professor has to be informed at least one week before the beginning of the examination session in order to provide students with the most suitable arrangements for each specific type of exam.