Artificial intelligence i

A.Y. 2020/2021
Overall hours
Learning objectives
The course aims to provide a general introduction to the Artificial Intelligence area considering the main and current sub-fields that characterize it. We will deal with classes of problems and associated solution methods at the basis of many of today's techniques that apply Artificial Intelligence in the real world.
Expected learning outcomes
The course will provide fundamental knowledge of the main areas of Artificial Intelligence, how they relate to the real world and to each other. The student will learn to recognize the problems with respect to which the techniques discussed in the course can provide a solution approach and to set their application. The ability to orientate oneself efficiently in the various areas of the discipline will be transmitted, providing a solid basis for targeted and autonomous in-depth studies.
Course syllabus and organization

Single session

Lesson period
First semester
For the theory part, lectures will be given synchronously from remote. A web conference platform (Teams or Zoom) will be used. Each lesson will be recorded and made available asynchronously through video files uploaded to the Ariel platform.

The programs will not be modified and the reference material will also include video lessons.

The written exam for the theory part will be done remotely ( platform or equivalent ones). Organizational details and any changes in the modalities will be communicated in time on the Ariel page of the course.

The modalities described above will be put in place in the event of restrictions imposed by the health emergency. They will be applied in whole or in part depending on the criticality of the specific situation. The organizational details will be communicated in time on the Ariel page of the course (
Course syllabus
The program is structured as follows:

- introduction to Artificial Intelligence, applications, research areas, and communities;
- autonomous and rational agents;
- automatic problem solving: formalization of the graph search problem, uninformed search, heuristic search;
- presence of adversaries: games and optimal strategies, game tree search;
- Constraint Satisfaction Problems: definition and resolution with search algorithms;
- uncertainty and sequential decisions: Markov Decision Processes;
- introduction to Machine Learning paradigms and basic approaches: decision trees, regression, and classification, non-parametric methods;
- neural networks: shallow and deep learning;
- Reinforcement Learning.

An advanced topic will also be presented in a concluding lecture, different in each year of the course, which describes a specific frontier problem of the field showing recent scientific contributions and open problems.
Prerequisites for admission
The course does not require any previous knowledge of the subject. However, for a better and easier understanding of the topics covered, basic knowledge of linear algebra, algorithms and probability and statistics are recommended. Having successfully attended the courses of Discrete mathematics, Algorithms and Data Structures, and Statistics and Data Analysis is a more than a sufficient guarantee.
Teaching methods
The theory part is given with frontal lectures where slides are presented. Slides are made available in PDF format through the Ariel platform ( Attendance is recommended.
Teaching Resources
The course is based on the topics presented in the book "Artificial Intelligence: A Modern Approach" by Peter Norvig and Stuart J. Russell.

For further support, slides and other supplementary material are provided during the course (
Assessment methods and Criteria
The exam consists of a written test lasting 2 hours where exercises and open questions with short answers are proposed.

The exercises require the application of the techniques discussed in class to problems of complexity appropriate to the duration of the test. Open questions assess knowledge of basic concepts and how they can be applied to problem-solving in the real world.

During the test, it is not allowed to consult any material.

The vote is out of thirty and will be communicated through the Ariel platform.

The assessments will take into account the mastery of techniques, correctness, and elegance of the solutions, clarity of presentation, knowledge of the concepts, and the ability to apply them in new settings. The exam and its evaluation will not be differentiated based on frequency.
INF/01 - INFORMATICS - University credits: 6
Lessons: 48 hours
Professor: Basilico Nicola
Educational website(s)