A class of gradient unconstrained minimization algorithms with adaptive stepsize (Q1970409)

The authors present a class of gradient algorithms with adaptive stepsize for the unconstrained minimization problem \[ f(x)\to \underset{x\in \mathbb{R}^n} {\text{minimum}}. \] The proposed class of algorithms comprises four algorithm: the first two incorporate techniques for the adaptation of a common stepsize for all coordinate directions and the other two allow an individual adaptive stepsize along each coordinate direction. Numerical tests are given.