A Spiking Self-Organizing Map Combining STDP, Oscillations, and Continuous Learning
Neural information coding is the fundamental of information processing mechanism in biological neural network. The study of neural information coding can help to understand the function of information processing in biological neural network and lay the theoretical foundation for improving bionic ability. As the abstract of a large number of real complex systems, small-world networks have the properties of biological neural networks. However, the neural information coding based on the small-world topology is rarely studied and the information transmission mechanism among the neurons is mostly excitatory regulation mechanism of spike-timing-dependent plasticity (STDP). In this paper, the small-world network is constructed and its properties are analyzed; the small-world spiking neural network based on the more complete STDP including excitatory synapse and inhibitory synapse is constructed; from the angle of firing rate of neurons and the temporal structure of the spike train, the properties of information coding on the small-world spiking neural network under the stimulations of white Gauss noise and impulse noise are analyzed respectively. Our experimental results indicate that under the same stimulation, the responses of the mean rate coding and ISI coding of the small-world network are both enhanced with the increase of stimulation intensity; under different stimulations, the mean rate coding and ISI coding of the small-world network show respective specificity.