Computing
CS423: Neural Networks (5 ECTS)
An introductory course in Neural Networks. Topics include learning algorithms, memory, the Rosenblatt perceptron, back-propagation multilayer perceptrons, and the Hopfield network.
Taught in Semester(s) 2. Examined in Semester(s) 2.
Workload: 99 hours (24 Lecture hours, 3 Lab hours, 72 Self study hours).
Module Learning Outcomes.
On successful completion of this module the learner should be able to:
- Describe the basic components of a neural network;
- Describe learning tasks for which neural networks are designed;
- Prove convergence of the Rosenblatt learning rule;
- Derive the weight update criteria for a multilayer perceptron;
- Calculate the optimal weight distribution for a Hopfield network.
Indicative Content
An introductory course in Neural Networks. Topics include learning algorithms, memory, the Rosenblatt perceptron, back-propagation multilayer perceptrons, and the Hopfield network.
Module Resources
- Simon Haykin, Neural Networks and Learning Machines, 3rd Edition, Prentice Hall.
Back