A Neural Network Approach for Solving Fractional-Order Model of HIV Infection of CD4+T-Cells

Abstract

In this paper the perceptron neural networks are applied to approximate the solution of Fractional-order model of HIV infection of CD4+T-cells that includes a system of fractional differential equations (FDEs). We converted this model to a system of Volterra integral equations. Then, by using perceptron neural networks ability in approximating a nonlinear function, we propose approximating functions to approach parameters of this system of Volterra integral equations. By obtaining the approximated solution of this system, the unknown parameters of the original fractional HIV model are adjusted. Numerical results illustrate this approach is simple and accurate when applied to systems of FDEs.