A New Spectral Conjugate Gradient Algorithm for Unconstrained Optimization Problems
Zhao Yue
Abstract
Zhao Yue
Abstract
A new spectral conjugate gradient method is proposed by combining Newton method and PRP spectral conjugate gradient method,which is a descent method.Furthermore,the new method is a combination of Birgin's spectral conjugate gradient method and PRP conjugate gradient method.The global convergence of the algorithm is proved under some assumptions.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
A new spectral conjugate gradient method is proposed by combining Newton method and PRP spectral conjugate gradient method,which is a descent method.Furthermore,the new method is a combination of Birgin's spectral conjugate gradient method and PRP conjugate gradient method.The global convergence of the algorithm is proved under some assumptions.
Key concepts: Conjugate gradient method, Derivation of the conjugate gradient method, Nonlinear conjugate gradient method, Conjugate residual method, Conjugate, Gradient descent, Gradient method, Convergence (economics)