07158 - Artificial Intelligence

Learning outcomes

At the end of the course, students:

• know the historical development and applications of artificial intelligence, defined both as the artificial simulation of the mind (emotions and cognitive processes) and of behavior (actions and decisions);

• know the most important differences differences, in terms of structure and potential applications, between artificial systems (e.g. classical connectionist algorithms, deep convolutional networks, role of expectations and prediction error);

• know the main aspects, for instance in terms of potential applications and of reception problems, linked to the introduction of artificial systems in real contexts (eg chatbots, self-driving cars, cognitive ergonomics);

• are able to link psychological concepts learned in the course of study, with their use in artificial systems aimed at detecting (e.g. affective computing, brain computer interfaces) or simulating (machine learning) mental states or human behavior.

Course contents

- Main steps in the history of artificial intelligence, from the first attempts to build "thinking machines" at the beginning of the 1900s, to the cybernetic approach of the 70s and to the subsequent popularity of connectionism over the years 1980s, up to the current deep neural networks.

- Application of the described algorithms to different areas (e.g. understanding of language, perception of images, prediction of behavior). The student will also know the main characteristics of artificial architectures based on rules, on heuristics, and on different types of learning (supervised, unsupervised, for reinforcement).

- Ability of artificial agents to interact with human beings (for example, social robotics) and will deepen the psychological, design and ethical implications of this area.

Readings/Bibliography

• Slides of the lessons
• Margaret A. Boden, L'intelligenza Artificiale, Il Mulino
• Paul Dumouchel, Luisa Damiano, Vivere con i robot, Raffaello Cortina Editore
• Articles and chapters indicated by the teacher during the course.

Teaching methods

Frontal lessons.

Assessment methods

The exam consistes in 40 multiple choice questions (4 alternatives). Time available: 30 minutes. The program includes all the bibliographic material indicated in the texts section, including any topic which was not discussed during the classes.

Each correct response has a score of 0.775, and the final result will be rounded up. “Lode” for tests with 39 or 40 correct responses.

Compensatory or dispensatory aids will be used, in the manner agreed with the teacher, for students diagnosed with SLD or other disability officially recognized and declared at the Service for students with disabilities and SLD of the University of Bologna.

Registration is required to take the exam through the function present on the AlmaEsami website, in compliance with the established deadlines.

Teaching tools

powerpoint presentations and IA examples on the web

Office hours

See the website of Andrea De Cesarei