Recurrent Approximation in the Tasks of the Neural Network Synthesis for the Control of Process of Phototherapy
Аlexander Trunov · River Publishers eBooks · 2022
The chapter reveals the problem of electromagnetic wave (EMW) interaction in the case when the electron of radical transition. The task of interaction between photon and electron with three natural frequencies different in three directions for terminal enzyme of respiratory chain cytochrome-c-oxidase or other photo acceptors was considered and solved. The solution of problems of encoding and decoding of polarized light interferogram data and calculation of radiation dose for different spectral composition are proposed. The examples of problem of analytical learning as standard elements for design multilayers artificial neural network (ANN) with any complicated configuration of input–output vector-function are considered. The vector-indicator was introduced and a new approximation–expansion within these three levels digital quantities was proposed. Recurrent approximation was applied for synthesis of recurrent artificial neural network (RANN). Solutions for synaptic weight coefficients (SWC) as recurrent sequence are represented in analytic form. Synthesis of control system as mini–max problem for efficiency optimization was solved. Convergence of recurrent approximation for two schemes of approach by a linear and quadratic curve are proposed and discussed for System Support Decision Making (SSDM) of physiotherapy with ANN and RANN.