[Comp-neuro] New paper: STDP enables spiking neurons to detect
hidden causes of their inputs.
nessler at igi.tu-graz.ac.at
Thu Nov 12 15:17:49 CET 2009
I would like to draw your attention to a new paper describing a
surprising theoretically founded connection between spike-timing
dependent plasticity (STDP) and the high-level mathematical concept
of unsupervised learning using expectation maximization (EM).
On the basis of this principle one can achieve unsupervised
learning results in networks of spiking neurons that were previously
The paper is available online at http://www.igi.tugraz.at/psfiles/191.pdf
B. Nessler, M. Pfeiffer, W. Maass.
STDP enables spiking neurons to detect hidden causes of their inputs.
In Proc. of NIPS 2009: Advances in Neural Information Processing Systems,
volume 22. MIT Press, 2010.
The principles by which spiking neurons contribute to the astounding
computational power of generic cortical microcircuits, and how
spike-timing-dependent plasticity (STDP) of synaptic weights could
generate and maintain this computational function, are unknown. We show
here that STDP, in conjunction with a stochastic soft winner-take-all
(WTA) circuit, induces spiking neurons to generate through their
synaptic weights implicit internal models for subclasses (or “causes”)
of the high-dimensional spike patterns of hundreds of pre-synaptic
neurons. Hence these neurons will fire after learning whenever the
current input best matches their internal model. The resulting
computational function of soft WTA circuits, a common network motif of
cortical microcircuits, could therefore be a drastic dimensionality
reduction of information streams, together with the autonomous creation
of internal models for the probability distributions of their input
patterns. We show that the autonomous generation and maintenance of this
computational function can be explained on the basis of rigorous
mathematical principles. In particular, we show that STDP is able to
approximate a stochastic online Expectation-Maximization (EM) algorithm
for modeling the input data. A corresponding result is shown for Hebbian
learning in artificial neural networks.
DI Bernhard Nessler
Institute for Theoretical Computer Science
Graz University of Technology
Inffeldgasse 16b, A-8010 Graz, Austria
nessler at igi.tugraz.at
Tel.: ++43 316 873-5823
Fax: ++43 316 873-5805
More information about the Comp-neuro