DOI: 10.36347/sjet

Pages: 529-534

In this paper, we propose and analyze a new non-monotone self-adaptive trust region method with fixed stepsize for unconstrained optimization. Unlike the traditional non-monotone trust region method, our algorithm utilizes a fixed formula to get the next iterative point if a trial step is not adopted. Besides, the trust region radius of related subproblem adjusts itself adaptively. By the above techniques, we can decrease the number of solving sub-problems efficiently. Under some standard assumptions, we show that the new proposed method has a global convergence.