Machine learning is concerned with the design of algorithms that can predict the evolution of a phenomenon based of a set of observations. Machine learning is a standard tool in the development of intelligent systems. It has been successfully applied to a wide range of domains, including vision, human-computer interaction, product recommendation, health, autonomous navigation, and many more. The course will describe and analyze, in a rigorous statistical framework, the most important machine learning techniques. This will provide the student with a rich set of methodological tools for understanding the general phenomenon of learning in machines.
Introduction Nearest Neighbour Tree predictors Cross validation Statistical risk Risk analysis for tree predictors Risk analysis for Nearest Neighbor Consistency Compression bounds Linear classifiers Online gradient descent From sequential risk to statistical risk Kernel functions Support Vector Machines Stability bounds and risk in SVM Boosting Neural networks and deep learning Dimensionality reduction
Calculus, discrete mathematics, statistics.
Prerequisiti e modalità di esame
The exam is written and oral (both parts are mandatory). The written part involves the preparation of a report describing either a set of theoretical results or the outcome of experiments on real datasets. The oral part includes a detailed discussion of the report and general questions on the course contents.
Exam; written and oral. Attendance: highly recommended. Teaching modality: traditional.
Materiale didattico e bibliografia
Shai Shalev-Shwartz e Shai Ben-David, Understanding Machine Learning: From Theory to Algorithms, Cambridge University Press, 2014.