A new method spectral conjugate gradient in unconstrained optimization

Abstract

In this study, we provide novel spectral conjugate gradient optimization techniques and analyze their convergence. Numerical tests reveal that our approaches can perform better than those already in use. the spectral conjugate gradient methods are commonly used for unconstrained optimization, especially when the dimension is large. Updated spectral techniques for tackling unconstrained optimization issues are developed based on curvature information. The strategies provided meet the descending criteria. The conjugate gradient algorithm is a powerful iterative method based on the parameters' conjugate gradient. We analyze the convergence properties of the algorithm and then give some numerical results which show the modified algorithms are robust and efficient. Furthermore, it is demonstrated that the innovative spectral techniques are globally convergent. The numerical findings show that the suggested techniques are successful when compared to the Fletcher –Reeves method.