Artificial Intelligence I

Artificial Intelligence I: problem solving methods, search techniques, propositional logic, knowledge representation Programming language: Prolog

Objectives

Upon successful completion of this course, the student will be able to: - Distinguish whether a problem requires the use of artificial intelligence methods , - Solve difficult combinatorial problems by searching the state space and satisfy the constraints - Solve logic problems with propositional logic, - Knowledge representation with complex data structures and ontologies, - Using Prolog programming language

Prerequisites

-

Syllabus

Introduction: Definition, history, relations with other scientific fields. Intelligent agents: rationality, performance metrics, work environments, structure of agents. Problem solving with search: State spaces, search trees, uninformed search methods (depth-first, breadth-first), search with partial information. Informed search and exploration: Best First and A* algorithms, local search algorithms (Hill climbing, simulated annealing, genetic algorithms). Constraint satisfaction problems: Constraint propagation, early check, arc consistency. Adversarial search: Optimal strategies in two-person games, minimax algorithm, alpha-beta pruning, extension to multiplayer games, extension to games of chance, expectiminimax algorithm. Logical agents: Propositional logic, inference patterns, resolution, logic circuits, first-order logic, inference rules for quantifiers, unification, inference chains, theorem proving, logic programming, introduction to Prolog language. Knowledge representation: Ontologies, representation of categories, objects, actions, states and occurences, semantic networks, description logics.