p. 174 — s ↦ f₂ᵢ(s, x₁(s), x₂(s)) is measurable, summable for i = 0 and essentially bounded for i ≥ 1
ProvedStochConvexProg.Duality.summable_of_boundedcarathodory-functionintegrabilityp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1stochastic-programming
Let the data satisfy the standing assumptions of the model, with a probability measure. Let and be bounded measurable functions. Then the functions
are measurable; the one with is summable (integrable), and those with are essentially bounded.
In particular the expected cost is a well-defined real number whenever is bounded and measurable, which is what makes the perturbation functional meaningful on .
Formalization Note Boundedness of is with respect to the sup norm of Fin n → ℝ; boundedness does not depend on the norm. Essential boundedness is an almost-everywhere bound.
Preamble
import Mathlib import Definitions.Def_StochConvexProg_Duality_Problem open MeasureTheory
Formal statement
namespace StochConvexProg.Duality
/-- Rockafellar–Wets (1976), p. 174: for bounded measurable `x₁ : S → Rⁿ¹`, `x₂ : S → Rⁿ²` the
functions `s ↦ f₂ᵢ(s, x₁(s), x₂(s))` are measurable, summable for `i = 0` and essentially bounded
for `i = 1, …, m₂`. -/
theorem summable_of_bounded {S : Type*} [MeasurableSpace S] {σ : Measure S}
[IsProbabilityMeasure σ] {n₁ n₂ m₁ m₂ : ℕ} (pr : Problem S σ n₁ n₂ m₁ m₂)
(x₁ : S → Fin n₁ → ℝ) (x₂ : S → Fin n₂ → ℝ)
(hx₁ : Measurable x₁) (hx₂ : Measurable x₂)
(hb₁ : ∃ B : ℝ, ∀ s, ‖x₁ s‖ ≤ B) (hb₂ : ∃ B : ℝ, ∀ s, ‖x₂ s‖ ≤ B) :
Measurable (fun s => pr.f₂₀ s (x₁ s) (x₂ s)) ∧
Integrable (fun s => pr.f₂₀ s (x₁ s) (x₂ s)) σ ∧
∀ i, Measurable (fun s => pr.f₂ i s (x₁ s) (x₂ s)) ∧
∃ B : ℝ, ∀ᵐ s ∂σ, |pr.f₂ i s (x₁ s) (x₂ s)| ≤ B := by sorry
end StochConvexProg.Duality
Source
Rockafellar and Wets, Stochastic convex programming: basic duality, Pacific J. Math. 62(1) (1976), p. 174, the sentence after the standing assumptions
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.