A Nonmonotone Line Search Technique and Its Application to Unconstrained Optimization II: R-Linear Convergence for Strongly Convex FunctionsResearch Paper
Line search methods for unconstrained minimization of a smooth function choose a direction and a step and set . Classical rules (Armijo, Wolfe) insist that every step decrease . For quasi-Newton and conjugate gradient directions this monotonicity requirement often forces short…