"Babes-Bolyai" University of Cluj-Napoca
Faculty of Mathematics and Computer Science

Implementation Techniques for Intelligent Systems
Code
Semes-
ter
Hours: C+S+L
Type
Section
MI098
8
2+0+2
optional
Informatica
MI098
8
2+0+2
optional
Matematică-Informatică
Teaching Staff in Charge
Assoc.Prof. SERBAN Gabriela, Ph.D.,  gabiscs.ubbcluj.ro
Aims
1. To present the field Intelligent Agents as a new research and development area in the field of Artificial Intelligence.
2. To present the main aspects related to the design and implementation of Intelligent Agents and how are they related to other programming paradigm (especially object oriented programming).
3. To present the field of reinforcement learning, a very actual and dinamic field of AI.
Content
1. Intelligent Systems
1.1. Artificial Intelligence: main aspects, research directions
1.2. Intelligent Agents: concept, structure, types, abstract and concrete architectures, agent oriented programming languages
1.3. The logic based representation and reasoning in agent based systemes

2. Searching
2.1 Uninformed searching
2.2 Informed searching (heuristics)

3. Game playing
3.1. Introduction
3.2. The MINIMAX search method
3.3. Alpha-beta pruning
3.4. Iterative deepening

4. Planning
4.1. Introduction
4.2. The Blocks World
4.3. The components of a planning system
4.4. Planning using stack of goals
4.5. Hierachical planning
4.6. Reactive systems
4.7. Other planning techniques

5. Learning
5.1. The general model of a learning agent
5.2. Learning a domain
5.3. Learning methods

6. Mathematical models for Intelligent Learning Agents
6.1. Markov Decision Processes
6.2. Partial Observable Markov Decision Processes
6.3. Hidden Markov Models

7. Reinforcement Learning
7.1. General concepts
7.2. RL based using an utility function
7.3. Q-learning
7.5. Action selection mechanisms (epsilon-Greedy, SoftMax)
7.6. The SARSA algorithm
References
1. SERBAN, G.:Sistememultiagent in Inteligenta Artificiala Distribuita. Arhitecturi si aplicatii, Ed. Risoprint, Cluj-Napoca, 2006
2. SERBAN, G., POP, HORIA F.:Tehnici de Inteligenta Artificiala. Abordari bazate pe Agenti Inteligenti, Ed. Mediamira, Cluj-Napoca, 2004.
3. POP, HORIA F. - SERBAN, GABRIELA: Inteligenta Artificiala. Cluj-Napoca: Centrul de Formare Continua si Invatamant la Distanta, 2003.
4. HARMON, M. - HARMON, S.: Reinforcement Learning - A Tutorial. Wright State University, 2000. [www-anw.cs.umass.edu./~mharmon/rltutorial/frames. html]
5. SUTTON, RICHARD S. - BARTO, ANDREW G.: Reinforcement learning. London : The MIT Press Cambridge, Massachusetts, 1998.
6. RUSSEL, STUART J. - NORVIG, PETER: Artificial Intelligence- A Modern Approach. New Jersey: Prentice- Hall, Inc., 1995.
7. FISHER, MICHAEL: Concurrent MetateM, A language for modeling reactive systems. Proceedings of Parallel Architectures and languages Europe (PARLE), Springer Verlag, 1993.
8. BROOKS, R. A.: A robot layered control system for a mobile robot. IEE Journal of Robotics and Automation, 2(1), 1986, pp.14-23.
9. WINSTON, P. H.: Artificial Intelligence. Addison Wesley, Reading, MA, 1984, 2nd ed.



Assessment
The final grade will be computed taking into account the following: lab activity (NA - 20%), a project realised during the semester(NP - 40%), the written exam (NE - 40%).
The access to the written paper is conditioned by the grade NP, which has to be at least 5. In order to promote the final exam, the final grade has to be at least 5.
Links: Syllabus for all subjects
Romanian version for this subject
Rtf format for this subject