Abstract:

Major advances in science often consist in discovering how macro?scale phenomena reduce to their microscale constituents. This article represents a brief, but practical introduction in the field of computational neuroscience. The astonishing hypothesis of computational neuroscience is that our minds can be explained by understanding the detailed behavior of neurons in the brain and their interactions with each other (Crick, 1994). Today, in this attempt, experimental and theoretical neuroscience studies are accompanied by mathematical theories of complex systems organization, evolutionary and developmental studies of the brain organization, all assisted by computational modeling means. First, the field of computational neuroscience is defined in relation with other research techniques of natural nervous systems. The cognitive modeling approach is motivated and we introduce the reader to a third generation of neural networks, based on computations with pulses. The level of detail in modeling the neural biophysics is considered next, and two classes of models are introduced and illustrated through examples. A general discussion on the computational properties of spiking neurons is aimed to point out the main advantages of pulsed neural networks over traditional networks.

Keywords: Computational neuroscience, spiking neural networks, detailed neural models, spike coding, spike response model.