Artificial Intelligence

Objectives

1. Adquirir conhecimento sobre a história da disciplina de Inteligência Artificial, os seus objetivos e a sua evolução.
2. Adquirir conhecimentos básicos ao nível dos fundamentos, enquadradas em formas de representação, de raciocínio e de aprendizagem automática.
3. Saber aplicar diferentes técnicas associadas à utilização de algoritmos explorando os diferentes temas abordados na disciplina.

Program

Introduction: A Brief history of Artificial Intelligence. Ethical implications and legislation.

Solving Search Problems: Blind and heuristic search. Search with adversary.

Knowledge and Uncertainty: Knowledge representation, propositional logic and knowledge engineering. Probability, Bayes' Rule and Bayesian Networks, Inference.

Optimisation: Concept of optimisation, Local search and meta-heuristics

Learning: Supervised, unsupervised and reinforcement learning

Neural Networks: Multilayer neural networks, gradient descent, backpropagation

Computer vision and convolution.

Teaching Methodologies

The theoretical sessions address problems that facilitate discussion of various topics throughout the semester. The problems focus on issues related to the topic covered. At the end of each session, students have the opportunity to sign up for Homework (TPC), selecting questions to be answered individually on Moodle and writing summaries on specific topics covered. The answers to the questions are always discussed in a later session, and the summaries produced are reviewed/corrected. In the laboratory sessions, students test the proposed algorithms, based on simple examples presented by the Professor. They develop group work following an experimentation guide. Ultimately, some groups may be eligible to sign up for work that could potentially lead to a presentation in the next session. In the project, students are challenged to choose an AI topic that interests them. This topic may address problems where algorithms are tested or other topics that address AI from a broader perspective (e.g. linking AI to education or ethical issues). All support materials are available on Moodle, which is also used to receive responses to the challenges proposed in the homework assignments, laboratory work and project work.

Bibliography

Russell S. e Norvig, P. (2020) Artificial Intelligence: A Modern Approach. Pearson, 4ª edição.

Costa, E. e Simões, A. (2008) Inteligência Artificial, Fundamentos e Aplicações. FCA. Domingos, P. (2017). A revolução do Algoritmo Mestre (2ª edição).