Eq. (3.2), p. 23 — the expected minimum of n independent uniforms on [0, 1] is 1/(n + 1)
ProvedRobustPower.CostGap.eq_3_2_expected_min_uniformorder-statisticsp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1probability
Let , let be a probability space and let be a measurable map whose coordinates are independent and each uniformly distributed on , that is, the law of under is the product of copies of the uniform distribution on . Then
This is the computational core of Theorem 3.1: together with Eq. (3.1) it gives . The paper calls (3.2) an inequality; it is the equality stated here.
Formalization Note The minimum of the coordinates is written as the infimum over the finite nonempty index set , which is attained; the hypothesis excludes the empty index set.
Preamble
import Mathlib open MeasureTheory
Formal statement
namespace RobustPower.CostGap
/-- Eq. (3.2), p. 23: if the coordinates of `d(ω)` are independent and uniform on `[0, 1]`, then
`E_μ[min(d₁(ω), …, dₙ(ω))] = 1/(n + 1)`. -/
theorem eq_3_2_expected_min_uniform {n : ℕ} (hn : 0 < n)
{Ω : Type*} [MeasurableSpace Ω] (μ : Measure Ω) [IsProbabilityMeasure μ]
(d : Ω → Fin n → ℝ) (hd : Measurable d)
(hlaw : μ.map d = Measure.pi (fun _ : Fin n => volume.restrict (Set.Icc (0 : ℝ) 1))) :
∫ ω, (⨅ j, d ω j) ∂μ = 1 / ((n : ℝ) + 1) := by sorry
end RobustPower.CostGap
Source
Bertsimas & Goyal, On the Power of Robust Solutions in Two-Stage Stochastic and Adaptive Optimization Problems, authors' manuscript (MIT DSpace) of Math. Oper. Res. DOI 10.1287/moor.1090.0440, p. 23, proof of Theorem 3.1, Eq. (3.2)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.