Proof of Lemma A.3, pp. 30, 35 — modulus from conjugate slopes
ProvedRegMT.Classif.lip_eq_sup_conjDomconjugateconvex-analysislipschitzp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1
For every convex Lipschitz continuous function , let and . Then
This identifies the exact coefficient in the final constraint of program (18). The Lipschitz modulus is the least Lipschitz bound, rather than any chosen bound.
Preamble
import Mathlib import Definitions.Def_RegMT_Classif_Model
Formal statement
namespace RegMT.Classif
/-- The Lipschitz modulus is the supremum of the absolute slopes in the
effective domain of the conjugate, pp. 30 and 35. -/
theorem lip_eq_sup_conjDom (L : ℝ → ℝ)
(hconv : ConvexOn ℝ Set.univ L)
(hLip : ∃ K : NNReal, LipschitzWith K L) :
WassersteinDRO.Duality.lipschitzModulus L = conjSlope L := by sorry
end RegMT.Classif
Source
Shafieezadeh-Abadeh, Kuhn & Mohajerin Esfahani, Regularization via Mass Transportation, arXiv:1710.10016v3, pp. 30, 35, end of proof of Lemma A.3 and proof of Theorem 3.11
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.