The Nonsmooth Stochastic Steffensen Type Methods and Their Applications to the Tensor Complementarity Problem
摘要
In this paper, we promote the stochastic Steffensen type method to solve a system of semismooth equations, and improve this algorithm from the perspectives of different selection criteria, minibatching, and Barzilai-Borwein step size. Finally, we propose several different modulus equations for the tensor complementarity problem and apply this series of nonsmooth stochastic Steffensen type methods to solve the tensor complementarity problem and the linear complementarity problem.