First- and Second-Order Conditions at a Penalised Extremum of a Test Function on the Wasserstein Space
lemmaAnalysisProbabilitylem:penalised-maximiser-wasserstein-2026aAt a point of the score domain at which a test function minus a positive multiple of the penalty has a local maximum on the penalty domain, the intrinsic gradient equals that multiple of the score and the translation Hessian is negative semidefinite; the mirror statement holds at a local minimum of the test function plus the penalty.
In the setting of Square-Integrable Vector Fields Against a Probability Measure on Euclidean Space, and Test Functions: Standing Notation, assume that is rich, let be a penalty pair on , let be positive and let be a test function on , with intrinsic gradient and translation Hessian at . The functions and on take the values and at , and local maxima and local minima relative to are taken in the metric space of Square-Integrable Vector Fields Against a Probability Measure on Euclidean Space, and Test Functions: Standing Notation §measures. For the score and the intrinsic gradient both belong to the tangent space , a subset of , in which and are real multiples. The set of symmetric real matrices, with the zero matrix , and its order are those fixed there. In this statement test function means a test function on and its intrinsic gradient; the test functions of Square-Integrable Vector Fields Against a Probability Measure on Euclidean Space, and Test Functions: Standing Notation §test keep their notation. Then the following hold.
1. (Penalised maximum)¶ Let be a point at which the function has a local maximum relative to . Then
2. (Penalised minimum)¶ Let be a point at which the function has a local minimum relative to . Then
Loading…
Prerequisites
No prerequisites tracked.
Dependents
No dependents yet.
Dependent proofs
No dependent proofs yet.
No relations recorded yet.