Proof of Theorem 2.4, display (p. 7) — upper-Lipschitz inverse (cited from [12], Prop. 2.1)
ProvedKingRockAsymp.Distribution.thm_2_4_display_upper_lipschitzgeneralized-equationsp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1set-valued-analysis
Let be a multifunction with , and suppose that the contingent derivative of at satisfies
Then there are a neighborhood of and a constant such that
for all sufficiently close to in , where is the closed unit ball of .
This is the upper-Lipschitz property of at . The paper cites it from King and Rockafellar's Sensitivity analysis for nonsmooth generalized equations ([12], Proposition 2.1) in the proof of Theorem 2.4; it turns the single-valuedness of the contingent derivative at into bounds in probability on solutions.
Formalization Note is the contingent derivative (2.2) of the inverse multifunction. "For all sufficiently close to " is ∀ᶠ y in 𝓝 0; and are chosen before .
Preamble
import Mathlib import Definitions.Def_KingRockAsymp_Distribution_Basic
Formal statement
namespace KingRockAsymp.Distribution
open Set Filter Topology Metric MeasureTheory ProbabilityTheory
theorem thm_2_4_display_upper_lipschitz {n m : ℕ} (F : Rn n → Set (Rn m)) (x₀ : Rn n)
(h0 : (0 : Rn m) ∈ F x₀)
(hD : contingentDeriv (svInv F) 0 x₀ 0 = {0}) :
∃ U ∈ 𝓝 x₀, ∃ lam : ℝ, 0 ≤ lam ∧
∀ᶠ y in 𝓝 (0 : Rn m), U ∩ svInv F y ⊆ closedBall x₀ (lam * ‖y‖) := by sorry
end KingRockAsymp.Distribution
Source
King & Rockafellar, Asymptotic Theory for Solutions in Statistical Estimation and Stochastic Programming, Math. Oper. Res. 18(1) (1993), proof of Theorem 2.4, display, p. 7 (authors' manuscript pagination); cited from [12], Proposition 2.1
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.