Training Feed Forward Neural Network with Modified Fletcher-Reeves Method
Yoksal A. Laylani, Khalil K. Abbo, Hisham Mohammed Khudhur · DergiPark (Istanbul University) · 2018
In this research, a modified Fletcher-Reeves(FR) conjugate gradient algorithm for training large scale feed forward neuralnetwork (FFNN) is presented. Under mild conditions, we establish that theproposed method satisfies the sufficient descent condition, and it is globallyconvergent under wolfe line search condition. The evidence which is provided byexperimental results showed that our proposed method is preferable and superiorto the classic methods.