A novel design of cascade-forward neural network for the chickenpox disease system

The aim of this work is to present the results of the chickenpox disease system numerically by relating an efficient optimization through a novel cascade-forward neural network. The chickenpox disease system is separated into classes of people: susceptible, infected without obstacles, infected, vaccinated, exposed, and recovered. The cascade-forward neural network is constructed by one hidden layer, optimized by Bayesian regularization, while the data is accessible through the Adams method. The data is obtained by using 0–1 intervals with the step size of 0.01, which moderates the mean square error with reasonable selection. This process contains 26 numbers of neurons, activation log-sigmoid function, one hidden layer and scale conjugate gradient process. The scheme’s exactness is approved through the matching of outputs, histogram error, insignificant absolute error, regression, and optimal validation performances. The proposed cascade-forward neural network is applied first to solve the chickenpox disease system.