ECE317 Artificial Intelligence

• Search for problem solving: definition of problems amenable to search, blind search algorithms (BFS, DFS, ID, B&B), heuristic search (BestFS, A*, HC), competitive search (Minimax).
• Knowledge Representation: Logic-based agent architecture, propositional logic, predicate logic and inference, rule-based systems and relation to logic programs, semantic networks, frames, conceptual graphs.
• Planning – deriving plans via searching state spaces, deriving plans via search in the plan space.
• Decision theory, preference theory, utility theory, decision under uncertainty, Markov decision processes. Uncertainty arising from the presence of other agents, game theory.
• Machine learning – decision trees, vector spaces.

The course aims to present the fundamental concepts and techniques of Artificial Intelligence and to highlight the philosophical problems that arise in the course of developing or using intelligent systems. The course addresses these issues from the perspective of intelligent agents, i.e., from a distributed artificial intelligence view, as this has become mainstream since 1995, and it brings together all technical and philosophical issues of interest. The main issues covered are included in the course content section. Students have the opportunity to gain knowledge in the main algorithms employed for problem solving, and in the main techniques involved in knowledge representation, as well as decision theory, reasoning under uncertainty and machine learning. They can identify the relative merits and disadvantages of the various techniques and apply them to a practical setting.