A Class of New Conjugate Gradient Method with the Wolfe Line Search
Zhou Hou-chun
Abstract
Zhou Hou-chun
Abstract
Conjugate gradient method is a method for solving unconstrained optimization problems,especially large-scale problems.In this paper a class of new conjugate gradient method contained the DY method is presented,and its global convergence is achieved under the Wolfe line search.Some numerical tests are carried out and the results show that the algorithm is effective.
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.
Conjugate gradient method is a method for solving unconstrained optimization problems,especially large-scale problems.In this paper a class of new conjugate gradient method contained the DY method is presented,and its global convergence is achieved under the Wolfe line search.Some numerical tests are carried out and the results show that the algorithm is effective.
Key concepts: Conjugate gradient method, Derivation of the conjugate gradient method, Conjugate residual method, Nonlinear conjugate gradient method, Conjugate, Convergence (economics), Gradient method, Biconjugate gradient method