Second-order characterizations of tilt stability with applications to nonlinear programming (Q2515034)

The notion of tilt stability has been extensively studied in the literature because of both theoretical and numerical reasons. In this paper a new approach to it is proposed in finite-dimensional spaces, which allows to derive not only qualitative but also quantitative characterizations of tilt-stable minimizers with calculating the corresponding moduli. Contrary to other contributions from the literature, this new approach and the main results derived by means of it hold, under appropriate modifications and using more involved tools of variational analysis, for optimization problems in infinite-dimensional spaces as well. Applications to classical nonlinear programs with twice continuously differentiable data are provided, too, a new Uniform Second-Order Sufficient Condition (USOSC) that is strictly weaker than the more conventional Strong Second-Order Sufficient Condition (SSOSC) being introduced and used for characterizing tilt stable local minimizers in nonlinear programming under additional qualification conditions. The paper is closed by brief concluding remarks, possible further developments and some open questions of the future research.