A radial basis neural network for the diarrhea disease model including treatment and vaccination
José Francisco Gómez-Aguilar, V. H. Olivares-Peregrino, Eduardo Pérez Careta, José R. Razo-Hernández, J. E. Lavín-Delgado, Jacinto Torres Jimenez · Advances in Complex Systems · 2026
This study aims to propose a stochastic radial basis (RB) artificial neural network (ANN) and the scale conjugate gradient (SCG) called RB-ANN-SCG for the diarrhea disease model, including treatment and vaccination (DDMTV). The diarrhea disease system is basically a susceptible, infected, and recovered model that includes the factors of treatment and vaccination. The dataset is obtained by using the Adam solver, which lessens the mean square error with the distribution of testing (12%), authentication (13%), and training (75%). A transfer RB function together with sixteen neurons is used in the hidden layers for solving the DDMTV, while the optimization is performed by the SCG. The scheme’s correctness is obtained via the overlapping of the reference and the achieved results. The insignificant calculated absolute error and best validation performances present the accuracy of the solver. Furthermore, the constancy and dependability of the RB-ANN-SCG are pragmatic through the histogram curves, function fitness, and correlation/regression for solving the DDMTV.