A New Approach to the Proximal Point Method: Convergence on General Riemannian Manifolds

作者：Bento Glaydston de Carvalho^*; da Cruz Neto Joao Xavier; Oliveira Paulo Roberto

来源：Journal of Optimization Theory and Applications, 2016, 168(3): 743-755.

DOI：10.1007/s10957-015-0861-2

• Proximal method
• Non-convex optimization
• Kurdyka-Lojasiewicz inequality
• Riemannian manifolds

In this paper, we present a new approach to the proximal point method in the Riemannian context. In particular, without requiring any restrictive assumptions about the sign of the sectional curvature of the manifold, we obtain full convergence for any bounded sequence generated by the proximal point method, in the case that the objective function satisfies the Kurdyka-Lojasiewicz inequality. In our approach, we extend the applicability of the proximal point method to be able to solve any problem that can be formulated as the minimizing of a definable function, such as one that is analytic, restricted to a compact manifold, on which the sign of the sectional curvature is not necessarily constant.

• 出版日期2016-3