Artificial intelligence: representation and reasoning [ LINGI2261 ]

> https://www.icampus.ucl.ac.be/claroline/course/index.php?cid=INGI2261

• Programminng abilities in a high-level language, algorithmics and data structures (e.g. SINF1121)
• Discrete mathematics (e.g. INGI1101)

• Problem solving by searching : formulating problems, uninformed and informed search search strategies, local search, evaluation of behavior and estimated cost, applications
• Constraint satisfaction : formulating problems as CSP, backtracking and constraint propagation, applications
• Games and adversarial search : minimax algorithm and Alpha-Beta pruning, applications
• Propositional logic : representing knowledge in PL, inference and reasoning, applications
• First-order logic : representing knowledge in FOL, inference and reasoning, forward and backward chaining, rule-based systems, applications
• Planning : languages of planning problems, search methods, planning graphs, hierarchical planning, extensions, applications
• AI, philosophy and ethics : "can machines act intelligently ?", "can machines really think ?", ethics and risks of AI, future of AI

Students completing successfully this course will be able to

• explain the basic knowledge representation, problem solving and reasonning methods in artificial intelligence
• assess the applicability, strength, and weaknesses of the basic knowledge representation, problem solving and reasonning in solving particular engineering problems
• develop intelligent systems by assembling solutions to concrete computational problems
• discuss the role of knowledge representation, problem solving and reasonning in intelligent-system engineering

Students will have developed skills and operational methodology. In particular, they have developed their ability to:

• master a new programming language using online tutorial
• deal with deadlines and competitivity in developping the most efficient solution.

• Exam : 75%
• Assignments : 25%.  Assignments 1,2,4,5 : we take the best three
  scores. Assignments must be personnal (team of 2). No collaboration between groups. No copying from Internet. Cheating = 0/20 all assignments

• Problem-Based Learning
• Learning by doing
• 5 assignments (one per two weeks)
• Team of two students
• Limited teaching (1 hour / week)
• Feed-back of problems (1/2 hour )
• Discussion of current problem (1/2 hour)

• Introduction
• Search
• Informed search
• Local search
• Adversarial search
• Constraint Satisfaction Problem
• Logical Agent
• First-order logic and Inference
• Classical Planning
• Planning in the real world
• Learning from examples
• Philosophical foundations & Present and future of AI

• Stuart Russell, Peter Norvig, Artificial Intelligence : a Modern Approach, 3nd Edition, 2010, 1132 pages, Prentice Hall
• slides online