1: \begin{abstract}A new type of stepsize, which was recently introduced by Liu et al. (Optimization 67(3):427-440, 2018), is called approximately optimal stepsize and is very efficient 	for gradient method. In this paper, we present an   efficient gradient method with approximately optimal stepsize  for large-scale unconstrained optimization, which combines the cubic regularization technique with a nonmonotone line search strategy. In the proposed method, if the objective function is not close to a quadratic on the line segment between the current and latest iterates, we construct a cubic regularization models to generate   approximately optimal stepsizes for gradient method. In particular, when the curvature is non-positive,  we   construct two special cubic regularization models to generate   approximately optimal stepsizes. Otherwise, we use quadratic models to generate   approximately optimal stepsizes for gradient method. A  modified strategy for  updating   the regularization parameter  is given.  The convergence of the proposed algorithm is established under the mild conditions. Numerical results indicated the proposed method is very promising.

2: 

3: \end{abstract}