Kourosh Parand - Department of Computer Sciences, Shahid Beheshti University, G.C., Tehran, Iran. Department of Cognitive Modelling, Institute for Cognitive and Brain Sciences,Shahid Beheshti University, G.C., Tehran, Iran
Amin Ghaderi - Department of Computer Sciences, Shahid Beheshti University, G.C., Tehran, Iran

In this article we suggest, the hybrid algorithm based on Bessel polynomials and Artificial Neural Network (BeNN) to solve non-linear Emden-Fowler type of differential equations.The problems with singular point at x=0 which have the second order of initial values, are modeled in mathematical physics and astrophysics. An unsupervised learning algorithm for a single layer Feed-Forward Neural Network model is proposed. The error back propagation is applied for updating the network parameters and minimizing the error functions. The hidden layer is eliminated by utilizing a series expansion of Bessel polynomials. The obtained outcomes demonstrate that the present methodology is applicable and reliable

Bessel artificial Neural Network, Emden–Fowler equation,Bessel polynomials, Error back Propagation method, Feed-forward Neural Network