Unit DECISION SUPPORT AND RECOMMENDER SYSTEM
- Course
- Informatics
- Study-unit Code
- GP004171
- Curriculum
- Intelligent and mobile computing
- Teacher
- Joseph Rinott
- CFU
- 9
- Course Regulation
- Coorte 2019
- Offered
- 2020/21
- Type of study-unit
- Obbligatorio (Required)
- Type of learning activities
- Attività formativa integrata
DECISION SUPPORT AND RECOMMENDER SYSTEM
Code | A001041 |
---|---|
CFU | 6 |
Teacher | Joseph Rinott |
Teachers |
|
Hours |
|
Learning activities | Affine/integrativa |
Area | Attività formative affini o integrative |
Academic discipline | MAT/06 |
Type of study-unit | Obbligatorio (Required) |
Language of instruction | English |
Contents | Elements of decision theory Elements of cooperative and not-cooperative games theory Elements of social choice theory |
Reference texts | K. Leyton-Brown, Y. Shoham: Essentials of Game Theory, Morgan & Claypool Publishers, 2008. D.M. Kreps: Notes On The Theory of Choice, Westview Press, 1988 W. Gaertner: A Primer in Social Choice Theory, Oxford University Press, 2009. Presh Talwalkar : The Joy of Game Theory: An Introduction to Strategic Thinking Paperback – 2014 Additional material will be provided by the teachers during the course. |
Educational objectives | The aim of the course is to acquire the main theoretical and methodological tools for modeling rational choices (both in presence of a single agent or for more agents) and to recommend the best choices for achieving the goals set. |
Prerequisites | The course requires basic knowledge of probability theory present in a first course of Probability and Statistics. All other knowledge required is covered by the undergraduate degree in Computer Science. |
Teaching methods | Frontal lessons that provide solutions of problems and cases study and exercises, usually held with the use of IT tools. |
Learning verification modality | The exam is a written test on all the topics taught in the course |
Extended program | DECISION THEORY AND GAME THEORY 1. Introduction to the course. Binary relations and their properties. Decision problem under certainty. Preference relations. 2. Probability background as needed. Subjective probability (Savage). Some background in statistics, and Bayesian statistics. 3. Lotteries, decisions under uncertainty. 4. Introduction to expected utility according to von Neumann-Morgenstern. von Neumann-Morgenstern axioms. 5. von Neumann-Morgenstern representation theorem. 6. Introduction to game theory. Various classifications of models in game theory. Examples, e.g. Prisoner’s Dilemma, the Chicken game, and their relation to current politics (Trump and North Korea?). Definition of non-cooperative strategic game. 7. Pareto optimality, best response, removal of dominate strategies, Nash equilibrium and its computation. 8. Nash equilibria in non-cooperative strategic games. Strictly competitive (or zero-sum) non-cooperative strategic games. Maxminimization, maxmin theorem, the relation to Nash equilibrium, and value of a strictly competitive (or zero-sum) game. Some examples. 9. Mixed and pure strategies for a non-cooperative strategic games. Expected utility for mixed strategy profiles. Mixed strategy Nash equilibrium. 10. Approximate Nash equilibrium, regret, Evolutionarily Stable Strategies (ESS). 11. Correlated equilibrium 12. Cooperative games and computation of Shapley’s value. 13. Games with sequential actions. 14. Repeated and stochastic games. 15. Statistics as a game and implications. Some discussion of statistical decision rules. 16. Paradoxes, Arrow’s impossibility theorem. 17. Social choice: aggregation of preferences, Gibbard Satterthwaite theorem, manipulations, majority rules and individual rights. |
DECISION SUPPORT AND RECOMMENDER SYSTEM
Code | A001052 |
---|---|
CFU | 3 |
Teacher | Marco Baioletti |
Teachers |
|
Hours |
|
Learning activities | Affine/integrativa |
Area | Attività formative affini o integrative |
Academic discipline | MAT/06 |
Type of study-unit | Obbligatorio (Required) |
Language of instruction | English |
Contents | Computational Intelligence: Neural Networks, Probabilistic Graphical Networks, Fuzzy Logic |
Reference texts | Computational Intelligence: An Introduction. Andries P. Engelbrecht. Second Edition Wiley 2007 Probabilistic Graphical Models Principles and Applications. Luis Enrique Sucar Springer 2015 |
Educational objectives | The aim of this course is to acquire the main concepts of Computational Intelligence and the ability of applying them to various problems in Artificial Intelligence |
Prerequisites | All knowledge required is covered by the undergraduate degree in Computer Science. |
Teaching methods | Frontal lessons that also provide solutions of problems and cases study and exercises |
Learning verification modality | The exam consists in an oral test (with duration of about 30 minutes) concerning all the concepts indicated in the program: more in detail, the student will be asked to describe some theoretical topics seen in the course. The purpose of this test is to ascertain the knowledge level, understanding capabilities and communication skills acquired by the student. Students who do not speak italian can do the exam in french or english. |
Extended program | 1. Neural Networks. Neurons and activation function. Feed-forward NN. NN Learning. Backpropagation. Gradient descent and other variants. 2. Recurrent networks. Convolutionary networks 3. Probabilistic model in AI 4. Probabilistic graphical models 5. Algorithms for inference and learning 6 Introduction to fuzzy logic and systems |