A Nonmonotone BFGS-TRUST-REGION Algorithm for Unconstrained Optimization Problems
Wu Qingjun
Abstract
Wu Qingjun
Abstract
In this paper,we propose a nonmonotone BFCS-trust-region algorithm which based on [19]for unconstained optimization.The advantage of this algorithm is that the subproblem of tust-region method is always a strictly convex quadratic programming. We prove the algorithm possesses global and Q-quadratic convergence under suitable conditions.
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.
In this paper,we propose a nonmonotone BFCS-trust-region algorithm which based on [19]for unconstained optimization.The advantage of this algorithm is that the subproblem of tust-region method is always a strictly convex quadratic programming. We prove the algorithm possesses global and Q-quadratic convergence under suitable conditions.
Key concepts: Trust region, Broyden–Fletcher–Goldfarb–Shanno algorithm, Convergence (economics), Quadratic programming, Mathematical optimization, Quadratic equation, Algorithm, Computer science