Advantages and Disadvantages of Spiking Neural Networks Compared to Classical Artificial Neural Networks

Spiking neural networks (SNNs) are a parallel branch of the development of artificial neural networks (ANNs) and SNNs are obviously significantly different from ANNs. In SNNs, unlike ANNs, information transmitted from neuron to neuron is an object that corresponds to a nerve impulse in the brain, in other words, a spike transmitted between neurons through synaptic communication. The functioning of the neuron in this case manifests itself only as the generation of spikes. Research shows that SNNs can work with single-bit signals with virtually no loss of accuracy. SNNs promise prospects that force such giants as IBM and Intel to create their research processors on the SNNs system. However, advantages also entail disadvantages - for example, there is still no final solution to the problem of how to train SNNs. The article reveals the advantages and disadvantages of spiking neural networks in comparison with classical artificial neural networks.