A New Formula for Conjugate Gradient in Unconstrained Optimization

The conjugate gradient method is an important part of the methods of optimization that are not constrained by local convergence characteristics. In this research, a new formula for the conjugated coefficient is derived depending on the linear structure. The new method fulfills the regression requirement. In addition, using the Wolff search line terms, the overall convergence of the new method has been demonstrated. At the end of the research were presented numerical results that show the effectiveness of the proposed method.