AUTHOR=Hayakawa Takashi , Kaneko Takeshi , Aoyagi Toshio 

TITLE=A biologically plausible learning rule for the Infomax on recurrent neural networks

JOURNAL=Frontiers in Computational Neuroscience

VOLUME=Volume 8 - 2014

YEAR=2014

URL=https://www.frontiersin.org/journals/computational-neuroscience/articles/10.3389/fncom.2014.00143

DOI=10.3389/fncom.2014.00143

ISSN=1662-5188

ABSTRACT=A fundamental issue in neuroscience is to understand how neuronal circuits in the cerebral cortex play their functional roles through their characteristic firing activity. Recently, several characteristics of cortical activity have been reproduced by Infomax learning of neural networks in computational studies. There are, however, still few models of the underlying learning mechanisms that allow cortical circuits to self-organize and display such characteristic activity. In the present article, we derive a biologically plausible learning rule for the maximization of information retained through time in dynamics of simple recurrent neural networks. Applying the derived learning rule in a numerical simulation, we reproduce several firing profiles observed in the cerebral cortex: cell-assembly-like repeats of precise firing sequences, neuronal avalanche, spontaneous replays of learned firing sequences and orientation selectivity in the primary visual cortex. We further discuss the similarity between the derived learning rule and the spike timing-dependent plasticity of cortical neurons.