Seminar Computational Intelligence B (708.112)

WS 2004/05

Institut für Grundlagen der Informationsverarbeitung (708)

Lecturer: O.Univ.-Prof. Dr. Wolfgang Maass

Office hours: by appointment (via e-mail)


Location: IGI-seminar room, Inffeldgasse 16b/I, 8010 Graz
Date: starting from Oct. 13, 2004 every Wednesday, 2.15 - 3.45 p.m.
(extra time slot fridays at 2.15 p.m. for make-up meetings in January)

Content of the seminar:

We will discuss theoretical methods that are needed for current research in the areas computational neuroscience and machine learning.


Schedule of talks:

Nov. 3: possibly Maribel (otherwise Maribel will speek on Nov 24)

Theoretical parts of some recent paper by Rajesh Rao
If he sends me sufficiently early a preprint of his paper in the next N IPS: Hierarchical Bayesian inference in networks of spiking neurons (Advances in NIPS 17, 2005, to appear), then this might be best. Otherwise the the talk will be about more theoretical parts of Bayesian computation in recurrent neural circuits (Neural Computation, 16(1), 1-38, 2004) or Predictive Coding in the Visual Cortex (Nature Neuroscience, 2(1), 79-87, 1999).

Nov. 24: Thuy
(if Maribel speeks on Nov. 24, then Thuy will move to Jan. 14)

Topic: Theory of  simulated annealing, probably from the book
Presentation: [PDF]

Dec. 1: Malte

Topic:  Theoritical results about Hebbian learning, STDP, and their relationship.
Suitable would be for example material from ch. 11 of the book by Gerstner and Kistler, or related papers by Gerstner et al.

Jan. 12: Stefan

Theory about interaction of excitation and inhibition in neural circuits, e.g. from N Fourcaud-Trocmé, D Hansel, C van Vreeswijk and N Brunel (2003), How spike generation mechanisms determine the neuronal response to fluctuating inputs, Journal of Neuroscience, 23, 11628-11640.
N Brunel (2000), Dynamics of sparsely connected networks of excitatory and inhibitory spiking neurons, Journal of Computational Neuroscience, 8, 183-208
Some basic theoretical arguments on that topics can also be found on pp. 265 of the book by Dayan and Abbot.

Jan. 14: Michael  (if Maribel speeks on Nov. 24 then Michael will move to Feb. 2)

Proofs of convergence for Reinforcement Learning Algorithms
(e.g. from ch. 6 of the book D. P. Bertsekas and J. N. Tsitsiklis. Neuro Dynamic Programming. Athena Scientific, Belmont, Massachusetts, 1996.)
Jan.19: Helmut

Balancing at the border of instability by L. Moreau  and Eduardo Sontag, Physical Review E 68(2003): 020901(1-4).
Online on
(There is also the opportunity to prove a related result which appears to be even more relevant for neural systems, where negative feedback aims at balancing some external excitation of a circuit).
Presentation: [PDF] [PPT]

Jan. 21: Joshi

Short term memory in echo state networks. GMD Report 152, German National Research Center for Information Technology, 2001 (60 pp.)

Jan. 26: Hr Bachler (on some paper related to feature selection) PDF

2004-10-05, by kkirch