A bio inspired learning scheme for the fractional order kidney function model with neural networks

The numerical procedures of the fractional order kidney function model (FO-KFM) are presented in this study. These derivatives are implemented to get the precise and accurate solutions of FO-KFM. The nonlinear form of KFM is separated into human (infected, susceptible, recovered) and the components of water (calcium, magnesium). Three cases of FO-KFM are numerically accessible using the stochastic computing scaled conjugate gradient neural networks (SCJGNNs). The statics assortment is performed to solve the FO-KFM, which is used as 78 % for verification and 11 % for both endorsement and training. The precision of SCJGNNs is achieved using the achieved and source outcomes. The reference solutions have been obtained by using the Adam numerical scheme. The competence, rationality, constancy is observed through the SCJGNNs accompanied by the imitations of state transition, regression performances, correlation, and error histograms measures.