Solving a class of biological hiv infection model of latently infected cells using heuristic approach

Abstract

The intension of the recent study is to solve a class of biological nonlinear HIV infection model of latently infected CD4+T cells using feed- forward artificial neural networks, optimized with global search method, i.e. particle swarm optimization (PSO) and quick local search method, i.e. interior- point algorithms (IPA). An unsupervised error function is made based on the differential equations and initial conditions of the HIV infection model repre- sented with latently infected CD4+T cells. For the correctness and reliability of the present scheme, comparison is made of the present results with the Adams numerical results. Moreover, statistical measures based on mean abso- lute deviation, Theil's inequality coefficient as well as root mean square error demonstrates the effectiveness, applicability and convergence of the designed scheme.