Numerical solution of third-order Emden–Fowler type equations using artificial neural network technique

Abstract

This paper presents the application of artificial neural network technique for solving a class of third-order linear and nonlinear boundary value problems with mixed nonlinear boundary conditions. This technique overcomes the singular behavior of problems and outlines the approximations of high exactness with a vast viable region of convergence. The proposed method has been tested for various examples; acquired outcomes exhibit the effectiveness and robustness of the proposed strategy and are compared with the other existing numerical techniques.