This module is an introduction to behavioural control systems. Specifically, it concentrates on properties of dynamical systems. This includes linear filters as well as nonlinear elements that are ordered serially, in parallel, or via recurrent connections. Following the understanding of such "simple" systems, the properties of massively parallel systems ("neural networks") will be considered. To develop a better intuition of the properties of such complex systems, software simulations will be performed, using customized programs, accompanied by theoretical instructions. An important aspect concerns training methods for artificial neural networks, thus introducing formal principles of plasticity and learning. These methods include supervised learning, unsupervised learning and learning with a critic. Practical application of the theoretical issues covered by this module will be based on modelling of biological systems. To this end, selected biological experiments will be performed and subsequently interpreted by means of computer simulations.
The entire workload (10 credit points, corresponding to 300 hours) will be rendered during the time of the module. Course time: 3.12.07 - 21.12.07 and 7.1.08 - 18.1.08. Preparation for examination: 21.1.08 - 25.1.08
Cruse, H. (2008) Neural networks as cybernetic systems. 3rd edition. Minds, Brains & Media.
ebook series; http://www.brains-minds-media.org/archive/1841
Rhythmus | Tag | Uhrzeit | Format / Ort | Zeitraum |
---|
Datum | Uhrzeit | Format / Raum | Kommentar zum Klausurtermin |
---|
Zeige vergangene Klausurtermine >>
Studiengang/-angebot | Gültigkeit | Variante | Untergliederung | Status | Sem. | LP | |
---|---|---|---|---|---|---|---|
Systems Biology of Brain and Behaviour / Master | (Einschreibung bis SoSe 2012) | Modul 2 | Pflicht | 1. | 10 | benotet |