Computational simulations of stochastic integrate-and-fire neurons balanced networks

Abstract

We propose to use computational simulations of integrate-and-fire neuronal network models to investigate their dynamical properties, such as the influence of network topology and synapse plasticity in their phase diagram and the nature of its critical regions. We intend to focus on neural networks that are balanced with respect to the ratio of excitatory and inhibitory inputs to each neuron. This balance seems to be fundamental for efficient learning and information processing by the network. We plan to investigate the phase diagram of such networks and better understand the homogeneous and heterogeneous regimes that have been observed in them. We also intend to investigate the role of synaptic plasticity in achieving said balance and in the appearance of self-organized criticality. (AU)