Artificial Intelligence 2016/2017
- 5 ECTS
- Taught in Portuguese
- Continuous Assessment
- relevant skillset
Students shall understand how intelligent systems are built, integrating solutions for real problems.
They shall understand the role of problem solving and learning techniques on the construction of intelligent systems.
Students shall be able to:
1. Formulate search problems in state spaces; understand the pluses and minuses of each search method, being able to select the best one for each problem.
2. Write programs to solve search problems in state spaces.
3. Identify, characterize and solve constraint satisfaction problems.
4. Characterize and solve simple multi-agent problems.
5. Extract knowledge from examples.
6. Understand neural network learning.
7. Apply statistical reasoning to learning and classification problems.
During theoretical-practical classes the subjects will be explained with practical examples and discussion.
During practical classes students will work on several assignments, having the opportunity to program the studied algorithms.
Body of Work
1. What is AI
2. Intelligent agents. Types of agents. Environments.
3. Problem solving by search. Problem formulation. Search strategies.
4. Non-informed and informed search.
5. Constraint satisfaction.
6. Games. Mini-max algorithm. Alfa-beta pruning.
7. Learning. Decision trees. Nearest neighbour classifier.
8. Neural networks.
9. Theorem of Bayes. Naive-Bayes classifier.
Stuart J. Russel and Peter Norvig. Artificial Intelligence - A Modern Approach, Pearson, Third edition, 2010, ISBN: 978-0-13-207148-2.
Ernesto Costa e Anabela Simões. Inteligência Artificial - Fundamentos e Aplicações, FCA, Segunda Edição, 2008, ISBN 978-972-722-340-4.
1. Introduction. History of AI. Agentes. Environments.
2. Types of agents. Search.
3. Non-informed search.
4. Informed search.
5. Problem resolution.
6. Constraint satisfaction.
7. Games. Mini-max algorithm. Alfa-beta pruning.
8. Learning. Nearest neighbour classifier
9. Test (date to be confirmed).
10. Decision trees.
11. Problem resolution.
12. Theorem of Bayes.
13. Naive-Bayes classifier.
14. Neural networks.
15. Test (date to be confirmed).
Demonstration of the syllabus coherence with the curricular unit's objectives
Topics 1 to 6 show the main techniques of problem solving by search and constraint satisfaction, and simple multi-agent systems (goal 1).
Topics 7 to 9 will show the students the basic mechanisms of machine learning and information extraction from large volumes of data (goal 2).
Demonstration of the teaching methodologies coherence with the curricular unit's objectives
Theoretical-practical classes allow the presentation, demonstration and discussion of the discipline subjects.
Practical classes will be used to solve concrete problems and follow the programming tasks execution.
|relevant generic skill||improved?||assessed?|
|Achieving practical application of theoretical knowledge||Yes||Yes|
|Analytical and synthetic skills||Yes||Yes|
|Balanced decision making||Yes||Yes|
|Commitment to effectiveness||Yes||Yes|
|Ethical and responsible behaviour|
|Event organization, planning and management||Yes|
|Information and learning management||Yes|
|IT and technology proficiency||Yes||Yes|
|Problem Analysis and Assessment||Yes||Yes|