Nada

2D1430, Artificiella neuronnät och neurofysiologi, 6 poäng


This is an introductory course, with the aim of giving basic knowledge about artificial neural networks (ANN) and the biological background to such models. The student will learn how neurons work and how they transmit information to other neurons in the network. He/she will also learn about the functional organization of the nervous system and the different computational tasks of the various substructures. Also the fundamental principles for how learning takes place in such networks will be discussed and related to learning rules used in the ANNs. The major types of network structures and learning rules will be treated, together with their different advantages and limitations, and their fields of applications. The student will learn how to model and simulate biologically realistic neural networks, as well as more abstract ANNs. Also models of cortical structures and dynamics will be dealt with, in particular with respect to perception and associative memory. There will be examples of technical applications of ANNs within areas like pattern recognition, combinatorial optimization, diagnosis etc. but also examples of hardware architectures for neural computation (neurochips and neurocomputers). All of these aspects will be covered in approximately twenty lectures and six computer labs.

Formell kursbeskrivning, det vill säga texten i studiehandboken.

Andra kurser som NADA ger för Främst D, E och F-elever.

Aktuell kursomgång: period 3-4 98/99

Aktuell information (senast ändrad 14 januari 1999) om den pågående kursen, kursledare Anders Lansner.

Tidigare kursomgångar och nästa kursomgång

^ Upp till kursöversikt.


Sidansvarig: <ala@nada.kth.se>
Senast ändrad 11 januari 1999
Tekniskt stöd: <webmaster@nada.kth.se>