On stochastic block methods for solving nonlinear equations
摘要
In this paper, we propose a double stochastic block method for solving nonlinear systems (DSBN), which unifies some stochastic row-block and column-block methods in a unified framework. Convergence analysis of the new algorithm is established based on local tangential cone conditions. Furthermore, to reduce the computations of the Jacobian matrix in random probabilities, a stochastic column-block method (SCBN) with uniform probabilities is established for solving nonlinear systems, and its convergence is investigated. Finally, numerical experiments are present to verify the effectiveness of the proposed methods.