Computational Neuroscience at Berlin School of Mind and Brain

Foundations of Computational Neuroscience: from cells to networks to organ

Get this page and list of lectures as word file.

Course description:

The course provides an introduction to selected basic concepts of computational neuroscience.

Qualification aims:

The participating students will get to know examples of models on several levels of abstraction, from single cells to networks and from basic concept, such as all-or-none response, to mental processes, such as learning and memory. The main goal is to teach specific fundamental computational principles that are often encountered in cortical processing.

Course content:

See complete list of all lectures.

This course intends to provide basic knowledge about computational models of excitable membranes, single cells, neuron populations, and neural networks, including some aspects of learning and neural coding, such as cortical feature maps. In addition, some basic aspects of dynamical systems and bifurcation theory will be taught, such as classification of excitability in type I and II, or tipping point behavior.

The course comprises:

• (a) Models of cells (membranes, Hodgking-Huxlex (voltage-gating), classification of excitability, integrate and fire)
• (b) Models of populations (rate- and activity based neural mass models, neural field models, oscillation, and synchronisation)
• (c) Models of networks (Encoding, perceptron, exclusive-or problem, Boolean function, state transition diagrams, hidden layer, error backpropagation ("credit assignment problem"))
• (c) Supervised, reinforcement, and unsupervised learning (self-organized map, learning by interacting with an environment)