1: \begin{abstract}

2: We investigate the inexact proximal point algorithm (PPA)  under  the bounded metric subregularity condition.  Global linear convergence of inexact PPA requires the knowledge of the bounded metric subregularity parameter, which is in general difficult to estimate. We propose an adaptive generalized proximal point algorithm (AGPPA), which adaptively updates the proximal regularization parameters based on a sequence of implementable criteria. We show that AGPPA achieves  linear convergence without any knowledge of the bounded metric subregularity parameter, and that the rate only differs from the optimal one by a logarithmic term. 

3:  We apply AGPPA to  convex minimization problem and analyze the iteration complexity bound. Our framework and the complexity results apply for arbitrary linearly convergent inner solvers, and are flexible enough to allow for a hybrid with any method. We illustrate the performance of AGPPA by applying it to  large-scale linear programming (LP) problem. The resulting complexity bound has weaker dependence on the Hoffman constant and scales with the dimension better than linearized ADMM. In numerical experiments, our algorithm demonstrates improved performance in obtaining solution of medium accuracy on large-scale LP problems.

4: \keywords{proximal point algorithm \and error bound condition \and  adaptive proximal regularization parameter \and  iteration complexity \and large-scale optimization \and linear programming}

5: % \PACS{PACS code1 \and PACS code2 \and more}

6: % \subclass{MSC code1 \and MSC code2 \and more}

7: \end{abstract}

8: