A few years ago Eugene Izhikevich introduced the concept of polychronization, a perspective on neural networks with conduction delays in which temporal input sequences generate signature-like temporal patterns in a neural network. In a follow-up article he proposes a solution for the distal reward problem. Here, a STDP-function is combined with a reward-function (representing Dopamine signaling) to strengthen certain connections of the recent past and thereby facilitating conditioning effects.

All publications come along with Matlab-code to reproduce the figures. I find the concept of polychronization very interesting but wonder whether there are further useful categories in which neural networks with various conduction delays can be characterized.

References

Izhikevich EM (2006) Polychronization: computation with spikes. Neural computation 18:245–82
Izhikevich EM (2007) Solving the distal reward problem through linkage of STDP and dopamine signaling. Cerebral cortex 17:2443–52

This is a restart of my weblog. From now on, everything here will be published in English and mostly deal with neuroscience and digital art.