A computational model of synaptic strength multiplicative scaling

Abstract:- In this paper we study the mathematical foundations of the phenomenon of multiplicative scaling of synaptic weights (strengths). The Hebbian learning rule that gave rise to the entire neural network area is only an approximation of what happens in Central Nervous System synapses. Conditional probabilities are postulated to match biological synaptic strengthening in a more realistic way. Multiplicative scaling of synaptic weights is a consequence of conditional probabilities calculation taking place at the level of synapses. Normalization of post-synaptic activity is also involved in the scaling (or normalization) of synapses