Nonmonotone Self-adaptive Trust Region Method and Its Convergence
Zhao Dan
Abstract
Zhao Dan
Abstract
The self-adaptive trust region method is an efficient tool to solve unconstrained optimal problems.In this paper,a self-adaptive trust region method with nonmonotone technique for unconstrained optimization problems was presented and analyzed.Under proper conditions,the global and local superlinear convergence rates were proved.Numerical results show that the new method is efficient.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
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.
The self-adaptive trust region method is an efficient tool to solve unconstrained optimal problems.In this paper,a self-adaptive trust region method with nonmonotone technique for unconstrained optimization problems was presented and analyzed.Under proper conditions,the global and local superlinear convergence rates were proved.Numerical results show that the new method is efficient.
Key concepts: Trust region, Convergence (economics), Mathematical optimization, Computer science, Rate of convergence, Mathematics, Key (lock), Economics