Lecture Component Artificial Intelligence
Hours 15 Lecture hours with 2 problem classes
1 Historical Overview - Definition of artificial intelligence (AI).Application areas. General problem solving versus specific knowledge. Complexity.
2-6 Heuristic Search - Uninformed versus informed search strategies. Formal properties of A*. Minimax game search, alpha-beta pruning.
7-10 Logic and Resolution - Knowledge representation. Propositional and predicate calculus. Inference rules. Clause form. Resolution strategies. Prolog and logic programming.
11-15 Uncertainty Reasoning - Probabilistic reasoning and Bayes theorem. Belief networks. Dempster-Shafer theory. Fuzzy logic.
Lecture Component Neural Networks and AI Programming
Hours 15 lecture hours with interspersed Problem Classes
1-2 Theory of logic programs - facts, queries, logical variables, recursion, rules, Horn clauses, structured data.
3-4 Basic Prolog - execution model, declarative and procedural meaning, backtracking, arithmetic, list representation, negation as failure and difficulties, simple examples.
5-7 Prolog Programming and Techniques - input/output, meta-logical and extra-logical predicates, set predicates, cuts, program development and style, correctness and completeness, Applications
8-10 Multi-Layer Perceptrons - Convergence theorem, non-separability, LMS algorithms, steepest
descent, back-propagation, generalisation, learning factors.
11 Radial Basis Function Networks - Multivariable interpolation, regularisation, comparison with
MLP, learning strategies.
12-13 Self-Organising Systems - Hebbian learning, competitive learning, SOFM, LVQ
14-15 Recurrent networks - energy functions, Hopfield net, nonlinear dynamical systems, Liapunov stability, attractors.