A generalization of artificial neuron by substitution of neural convolution to arbitrary multi-parametric function. Weights of synapses have been changed to the set of parameters and a learning algorithm has been modified as well. Even one single “smart neuron” is sufficient for finding...

Based on artificial intelligence approach it's possible to show a valuable link between phenomena of periodical activity/sleeping phases for living beings and implementation/training phase for artificial neural network acting in real time . In this article also reveals a deep link of another...