On the best choice of a damping sequence in iterative optimization methods (Q1111477)

Some iterative methods of mathematical programming use a damping sequence \(\{\alpha_ t\}\) such that \(0\leq \alpha_ t\leq 1\) for all t, \(\alpha_ t\to 0\) as \(t\to \infty\), and \(\sum \alpha_ t=\infty\). For example, \(\alpha_ t=1/(t+1)\) in Brown's method for solving matrix games. In this paper, for a model class of iterative methods, the convergence rate for any damping sequence \(\{\alpha_ t\}\) depending only on time t is computed. This computation is used to find the best damping sequence.