+++ This website was last updated on Feb 17, 2014 +++
DAY ONE
01) Introduction to computational neuroscience / Mo. 9:15 - 10:45
In which we will take a broad view on the subject, talks about models, and identify our objectives of the course.
- Slides
02) Hodgkin-Huxley model (conductance-based membrane model) / Mo 11:00 -12:30
In which we will derive from scratch of the most successful model in mathematical biology.
- My script for the white board.
03) Geometry of excitability / Mo 13:30 - 15:00
In which we understand that information processing depends also on electrophysiological properties and not exclusive on synaptic circuits.
-My script for the white board.
04) Action potential propagation propagation / Mo 15:15 -16:45
In which the temporal dynamics spread their wings and fly through space.
-My script for the white board.
- Link to the paper from Karl Friedrich Bonhöfer "Modelle der Nervenerregung" (yes, it's in German).
DAY TWO
05) Models of neural populations (rate-based) / Tue 9:15 - 10:45
In which we will asked: what is the rate? The average membrane potential, the mean firing rate, or the fraction of active cell in the population. And don't get an answer.
-My script for the white board
recommended paper in connection also with 08
- Dynamics of non-convulsive epileptic phenomena modeled by a bistable neuronal network.
06) Oscillations, Synchronization, and tipping points / Tue 11:00 -12:30
In which we apply the power of mathematics to the dynamics of romance; we'll learn that most of us are a 'cautious lovers' for a reason.
recommended paper
- Synchronization of Cellular Clocks in the Suprachiasmatic Nucleus
07) 2nd-generation Hodgkin-Huxley model (ion-based tissue model) / Tue 13:30 - 15:00
In which we gain a greater depth of knowledge on the Hodgkin-Huxley formalism.
08) Dynamical diseases of the brain / Tue 15:15 -16:45
In which you'll see me following Humboldt's ideal of the coexistence of research and teaching; yes I'll make my life easy in this lecture.
Wednesday we will rest and go to the talk form Steve Schiff.
DAY THREE
09) Neural code, decoding, topographic sensory input representation / Thu 9:15 - 10:45
In which we'll see an example of how a model of data suddenly turns into a model of phenomena.
-My script for the white board.
10) Perceptron and learning rules / Thu 11:00 -12:30
In which we will disenchant the perceptron and talk about matrix operations and the uplifting bias.
-My script for the white board.
11) Networks, N-cubes, and cycles in state transition diagrams / Thu 13:30 - 15:00
In which salamanders and leeches walk like an egyptian due to asynchronous updates in a central pattern generator.
-My script for the white board.
recommended paper in connection also with 08
Structure and dynamics of neural network oscillators
12) Hidden layer, credit assignment problem, and backpropagation / Thu 15:15 -16:45
In which we'll look back to Helmholtz efference copy and will learn how a model learns the reafference principle (spoiler: it's by the chain rule we know from calculus).
DAY FOUR
13) Self-organizing maps / Fr 9:15 - 10:45
In which we will think about where to place what in the shelves of a grocery store based on shopping data that's collected via loyalty cards.
14) Hopfield network / Fr 11:00 -12:30
In which we will learn again (cf. backpropagation) about "landscapes" (previously, we will have considered error gradient descent), now we use an energy concept to explain associative memory.
15) Recapitulation / Fr 13:30 - 15:00
In which you let me know what you have not understood, so we can go over it again.
16) Multiple choice test / Fr 15:15 -16:45
In which I will learn what you have learned. 30 questions are waiting for you.