#### Learning outcomes

On successful completion of this module, students will be able to:

1. appreciate the range of applicability of intelligent learning concepts and techniques;

2. explain the state of the art of intelligent learning concepts and techniques;

3. have sufficient knowledge of a basic rule induction method, such as ID3, to be able to explain it;

4. explain the theory behind artificial neural networks, in particular perceptron and back propagation;

5. describe and explain the theory behind concept learning, Baysian learning, genetic algorithm, reinforcement learning, and inductive logic programming;

6. apply Baysian theory to justify concept learning, decision tree learning, artificial neural network learning.

#### Brief description

This module builds on CS26210 and examines some of the key ideas in Artificial Intelligence. A small number of topics are studied in depth in order to give insight and understanding of the methods and issues involved in state-of-the-art developments.

#### Content

1. Introduction - 2 lectures

Possibility and necessity of learning, target function,components in a learning system, performance measurement of learning systems.

2. Concept learning - 3 lectures

Generality ordering of hypotheses, FIND-S algorithm, candidate elimination algorithm, version space, the LIST-THEN-ELIMINATE algorithm, inductive bias

3. Decision tree learning - 3 lectures

Entropy, best attribute, information gain, best tree, inductive bias, Occam's razor, over-fitting, reduced error pruning, rule post-pruning

4. Artificial neural network - 3 lectures

Perceptron, linear separability, gradient decent, sigmoid function, back propagation algorithm, over-fitting

5. Baysian learning - 4 lectures

Baysian theory, maximum a posteriori hypothesis, maximum likelihood, probability density, normal distribution, minimum description length principle, Bayes optimal classifier, naive Bayes classifier

6. Genetic algorithm - 3 lectures

Best hypothesis, hypothesis representation, genetic operators, fitness function, fitness proportionate selection, steady state selection, rank based selection

7. Reinforcement learning - 2 lectures

Q-learning, adaptive dynamic programming, temporal difference learning.

#### Reading Lists

**Books**

**** Recommended Text**

S J Russell and P Norvig (1995) *A I: A Modern Approach *
Prentice-Hall 0131038052

S. M. Weiss and C. A. Kulikowsky (1991) *Computer Systems That Learn *
Morgan Kaufmann

T. Mitchell (1998) *Machine Learning *
McGraw Hill

#### Notes

This module is at CQFW Level 6