Neural networks

Neural networks refer to dynamical properties of interconnected units (neurons). From a computational point of view, it is possible to distinguish two different subfields. Artificial neural networks is a branch of Artificial Intelligence defined as "...a computing system made up of a number of simple, highly interconnected processing elements, which process information by their dynamic state response to external inputs". On the contrary, biological neural networks are computational implementations of anatomical circuitry used to investigate brain properties. The study of this latter type has implications on a broad spectrum of fields in brain sciences.

Introduction
A biological neural network is an interconnected web of neurons transmitting elaborate patterns of electrical signals. The goal of a biological neural network is to simulate electrophysiological dynamics of modules of neurons. Such results are achieved by modeling cellular dynamics of individual neurons and by connecting such units following some biological criteria.

Neurons and Connectivity
In this kind of simulations it is possible to reproduce spiking behavior of a neuron. Each unit is characterized by a pre-synaptic input (dendrite) where upstream neurons send information. The summation of such incoming potential determines the occurrence of an action potential in the neuron. In case of depolarization (membrane's voltage tends towards equilibrium), the cell will emit a spike and pass information further on to downstream units via an axon and its synapses. The collective behavior of modules of neurons gives rise to a variety of activity dynamics.

Models
Neural networks comprises a multitude of neuron and connection models to simulate experimental recordings of both in vivo and in vitro modules. Examples of single neuron models include Hodgkin–Huxley, Leaky Integrate-and-Fire (LIF), Izhikevich and adaptive exponential LIF by Brette and Gerstner. In particular, in this project the LIF model has been employed as it provides a straightforward approach.

Software
Many simulating software have been developed in the last decades in order to facilitate computational implementations of such anatomical structures. Among the most popular in the research community are NEURON, NEST , Brian2. The present wiki deals with models implemented in the latter software.