MODIFIED LIMITED MEMORY BFGS METHOD WITH NONMONOTONE LINE SEARCH FOR UNCONSTRAINED OPTIMIZATION
Gonglin Yuan, Zengxin Wei, Yanlin Wu
Abstract
Open-access reader
Gonglin Yuan, Zengxin Wei, Yanlin Wu
Abstract
Open-access reader
In this paper, we propose two limited memory BFGS algorithms with a nonmonotone line search technique for unconstrained optimization problems. The global convergence of the given methods will be established under suitable conditions. Numerical results show that the presented algorithms are more competitive than the normal BFGS method.
OpenAlex reports 29 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.
In this paper, we propose two limited memory BFGS algorithms with a nonmonotone line search technique for unconstrained optimization problems. The global convergence of the given methods will be established under suitable conditions. Numerical results show that the presented algorithms are more competitive than the normal BFGS method.
Key concepts: Broyden–Fletcher–Goldfarb–Shanno algorithm, Line search, Convergence (economics), Mathematical optimization, Mathematics, Line (geometry), Algorithm, Computer science