The Discrete Neuronal Model and the Probabilistic Discrete Neuronal Model

In this paper, we present and analyze a class of neural network models previously proposed by the author: the discrete neuronal model and the probabilistic discrete neuronal models (DN and PDN). The models have been applied to model some biological neural networks (e.g. the stomatogastric ganglion in lobsters) and domenstrated their generality and usefulness in simulating the information processing capabilities of real neural networks given their exact or inexact connectivity patterns and endogenous firing patterns. Another application of this model is in building rule based inference systems. The advantages of this type of models is that it can handle variable bindings easily. We also incorporate certainty factors propagation mechanisms into the system. Systems for sequential processing based on this models have been worked on too.