A Bayesian regularization intelligent computing scheme for the fractional dengue virus model

This research’s goal is to investigate the numerical assessments of a fractional order dengue viral model (FO-DVM) by using the artificial intelligence procedure of Bayesian regularization neural networks (BRNNs). The FO derivatives present more precise results as compared to integer order for solving the DVM. The dynamics of the mathematical DVM form is considered into five classes. The computing stochastic BRNNs approach is presented for three variations with the selection of the data as testing 13%, authentication 11% and training 76% together with sixteen hidden neurons. The result’s comparison is accessible in the form of overlapping, which is based on the BRNNs approach and reference Adam solutions. However, minor absolute error around 10^-05 to 10^-07 enhances the worth of the proposed solver. The BRNNs approach is used to minimize the mean square error for the mathematical FO-DVM. The obtained measurements of error histograms values, and regression coefficient calculated as 1 are presented to verify the efficiency of stochastic BRNNs approach.