Artificial Neural Network For Solving System of Nonlinear Ordinary Differential Equations

Abstract

In this thesis we implemented Artificial neural network for solving system of non linear ordinary differential equations. We develop a vectorized algorithm for solving SEIR(S-Susceptible ,E-Exposed, I-Infectious and R-Recovered) model and apply python code.We develop more techniques to handle the challenges of experiments. For the learning of the neural network, we utilized the adaptive moment minimization method. Finally,we compare with Runge-Kutta order four method and We have shown that, the artificial neural network could gives better accuracy.