Proposition 1 (i), p. 6 — z*_n → z*, w.p.1
ProvedSolutionQuality.SRP.prop1_iconsistencyp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1sample-average-approximationstochastic-programming
Assume (A1)–(A3), and let be i.i.d. as . Let be the optimal value of (SP) and the optimal value of the sample average approximation (SP). Then
The optimal value of the sampled problem is a strongly consistent estimator of the true optimal value; this is part (i) of Proposition 1, used in the proof of part (ii).
Formalization Note. The paper's hypothesis is not used by this part and is dropped (the statement is stronger without it). is assumed measurable for each . and are infima of images of ; under (A1)–(A3) they are attained on the probability-one event where the sample objective is continuous.
Preamble
import Mathlib import Definitions.Def_SolutionQuality_SRP_Setting
Formal statement
namespace SolutionQuality.SRP
open MeasureTheory ProbabilityTheory Filter Topology
theorem prop1_i {d : ℕ} {Ξ : Type*} [MeasurableSpace Ξ]
(μ : Measure Ξ) [IsProbabilityMeasure μ]
(f : E d → Ξ → ℝ) (hfm : ∀ x, Measurable (f x))
(X : Set (E d)) (hA : Assumptions μ f X)
{Ω : Type*} [MeasurableSpace Ω] (P : Measure Ω) [IsProbabilityMeasure P]
(ξ : ℕ → Ω → Ξ) (hξ : IsIIDSample P μ ξ) :
∀ᵐ ω ∂P, Tendsto (fun n => saaValue f X (fun i => ξ i ω) n) atTop
(𝓝 (optValue μ f X)) := by sorry
end SolutionQuality.SRP
Source
Bayraksan & Morton, Assessing Solution Quality in Stochastic Programs, preprint (January 26, 2005), p. 6, Proposition 1 (i)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.