pike Response Function (SRF) plays an important role in the temporal coding Spiking Neural Network (SNN) as it has a significant role to determine when the neuron should fire. This paper studies the important role of the SRF in the SNN learning stability. It proposes a novel method to find out the rules to update delay for each class to make SRF stable, and then using these rules to update delay and weight simultaneously at the SNN learning rule. This method updates the delay depending on the local result to make SRF stable. The main issue of this paper is to put forward the idea that weight and delay parameters could and need to be updated simultaneously to make both SRF and SNN stable during the learning process. The delay rules strategy which have been found could be used for pattern recognition application which use SNN. The limitation of this work is that; getting the updating delay rules depends on a sample data from each class and the way of selecting the rules.
|Number of pages||5|
|Journal||International Journal of Engineering and Technology|
|Publication status||Published - 26 Nov 2018|