About the Mechanism of Neural Network as Adaptive System Training
Lorents V.A., Gavrikov V.L., Khlebopros R.G.
FGAOU VPO Siberian Federal University,
V.P.Astafev Krasnoyarsk State Pedagogical University,
Krasnoyarsk Scientific Center of the Russian Academy of Sciences
Abstract. The dynamics of error learning neural networks, obtained under identical conditions of the problem being solved and the parameters of networks, but with different initial conditions of the network are analyzed. And also a possible theoretical model of a single mechanism for training of artificial and natural adaptive systems is attempted to offer.
Key words: neural networks, learning, training error.