(Winter term; Block course during 1 week; 15 hours total )

This course is intendend for graduate students in the Neurosciences with an interest in Computational Neuroscience. The contents are similar to that of the course `Neural Networks and Biological Modeling'. However, since the course is aimed for experimentalists in neuroscience, some mathematically more difficult chapters have been removed. Nevertheless an elementary knowledge of differential equations is required.
See also the page of the Doctoral School in Neuroscience.

Objectifs:
Be introduced to the basic questions and models in Computational Neurosciences: single-cell modeling, Hebbian unsupervised learning, reward-based learning, models of associative memory, population dynamics, and spike-timing dependent synaptic plasticity.

Contents

• Lecture 1: Neural Coding and Integrate-and-fire model
• Lecture 2: Hodgkin-Huxley equations and detailed neuron models
• Lecture 3: FitzHugh-Nagumo and analysis of 2-dimensional neuron models (not given in 2006)
• Lecture 4: Synaptic plasticity and unsupervised learning
• Lecture 5: Associative memory and population dynamics
• Lecture 6: Rewared-based learning and behavior

  Exam: Oral exam.

  Recommanded texts

  • For topics related to spiking neurons, look at Spiking Neuron Models by W. Gerstner and W. Kistler (Cambridge Univ. Press). In particular chapters 1-4 and 10.
  • As a background text, look at Theoretical Neuroscience by P. Dayan and L.F. Abbott (MIT Press).