Corollary A2 — the empirical mean is a -valued random variable
ProvedKingRockAsymp.Distribution.corollary_A2measurabilityp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1probability
Let be compact, let and the samples satisfy P.1–P.4, and let each be measurable. Then for every positive the empirical mean
is a random variable with values in , i.e. a Borel measurable map from the sample space into .
This is what makes the convergence in distribution of in Theorem A3 meaningful.
Formalization Note The sum runs over (Finset.range ν), i.e. is s i. The corollary requires , where the paper defines the sample mean. P.1–P.4 are inherited from the Appendix's standing assumptions.
Preamble
import Mathlib import Definitions.Def_KingRockAsymp_Distribution_Basic
Formal statement
namespace KingRockAsymp.Distribution
open Set Filter Topology Metric MeasureTheory ProbabilityTheory
theorem corollary_A2 {n m : ℕ} {S Ω : Type*} [MeasurableSpace S] [MeasurableSpace Ω]
(P : Measure Ω) [IsProbabilityMeasure P]
(U : Set (Rn n)) [CompactSpace ↥U] (f : S → C(↥U, Rn m))
(hmeas : ∀ x : ↥U, Measurable (fun σ => f σ x))
(s : ℕ → Ω → S) (hs : ∀ i, Measurable (s i))
(hindep : iIndepFun s P) (hident : ∀ i, IdentDistrib (s i) (s 0) P P)
(hP3 : ∃ x : ↥U, MemLp (fun ω => f (s 0 ω) x) 2 P)
(hP4 : ∃ a : S → ℝ, MemLp (fun ω => a (s 0 ω)) 2 P ∧
∀ σ : S, ∀ x₁ x₂ : ↥U, ‖f σ x₁ - f σ x₂‖ ≤ a σ * ‖(x₁ : Rn n) - x₂‖) :
∀ ν : ℕ, 1 ≤ ν →
Measurable (fun ω => (ν : ℝ)⁻¹ • ∑ i ∈ Finset.range ν, f (s i ω)) := by sorry
end KingRockAsymp.Distribution
Source
King & Rockafellar, Asymptotic Theory for Solutions in Statistical Estimation and Stochastic Programming, Math. Oper. Res. 18(1) (1993), Appendix, Corollary A2, p. 16 (authors' manuscript pagination)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.