Study of the Influence of Memristive Plasticity on the Formation of a Feature Space in the Excitatory Layer of a Spiking Neural Network

In this study, we present a novel image decoding model utilizing a spiking neural network enhanced with memristive plasticity, tailored for recognition tasks. We introduce the average number of spikes in the excitatory neuron layer as a distinctive feature space. Through extensive experimentation with digital images from the MNIST database, we demonstrate that our model achieves efficient feature separation by incorporating spike timing-dependent memristive plasticity in excitatory synapses.