Evolving Spiking Neural Network (ESNN) and Harmony Search Algorithm (HSA) for parameter optimization

Spiking Neural Network (SNN) acts as a part of the third generation of Artificial Neural Networks (ANNs). Evolving Spiking Neural Network (ESNN) is one of the most broadly utilized among in SNN models in numerous current research works. During the classification process, ESNN model acts as a classifier and three parameters are used in this article. However, the parameters are needed to set manually before the classification process. To solve the stated problems, ESNN required an optimizer that able to optimize the three parameters such as similarity value, modulation factor and proportion factor. The best estimations of parameters are adaptively chosen by Harmony Search Algorithm (HSA) to abstain from choosing appropriate values for specific issues through the trial-and-error approach. Therefore, this article proposed the integration of ESNN as a classifier and HSA as an optimizer for parameter optimization. The experimental results give favorable accuracy rates via the hybrid of ESNN and HSA.