Proof of Theorem 2.1, p. 6, case "only random costs" (r = 1, r′ = ∞) — v(ξ̃) − v(ξ) ≤ L(‖ξ̃ − ξ‖_1 + Σ‖x̄_τ − E[x̄_τ|F̃_τ]‖_∞) + ε
ProvedMultistageStab.Value.case_costs_randommultistagep2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1stabilitystochastic-programming
Work in the setting of program (1) when only the costs are random ( deterministic, , ). Assume (A1), bounded, admissible, finite, and the level-boundedness of (A2) with constants , . Then there is a constant such that for every , every admissible with and finite, and every ,
This is the second case estimate in the proof of Theorem 2.1; here the decisions are essentially bounded.
Formalization Note As in the other case estimates, the inequality is stated in with its left-hand side truncated at , and is the essential supremum norm (eLpNorm … ⊤).
Preamble
import Mathlib import Definitions.Def_MultistageStab_Value_Setting open MeasureTheory ProbabilityTheory open scoped ENNReal
Formal statement
namespace MultistageStab.Value
theorem case_costs_random (D : Data) (hpat : Pattern.costs.Holds D)
{Ω : Type*} [MeasurableSpace Ω] (P : Measure Ω) [IsProbabilityMeasure P]
(ξ : ℕ → Ω → E D.d) (hA1 : CompleteRecourse D)
(hv_bot : value P D .costs ξ ≠ ⊥) (hv_top : value P D .costs ξ ≠ ⊤)
(α δ : ℝ) (R : NNReal) (hα : 0 < α) (hδ : 0 < δ)
(hA2 : LevelBoundedWith P D .costs ξ α δ R)
(hA3 : Admissible P D .costs ξ) (hX1 : Bornology.IsBounded (X1 D)) :
∃ L : ℝ, 0 < L ∧ ∀ ε : ℝ, 0 < ε → ε ≤ α →
∀ ξ' : ℕ → Ω → E D.d, Admissible P D .costs ξ' →
inputDist P D .costs ξ ξ' < ENNReal.ofReal δ →
value P D .costs ξ' ≠ ⊥ → value P D .costs ξ' ≠ ⊤ →
∀ xbar ∈ levelSet P D .costs ξ ((value P D .costs ξ).toReal + ε),
ENNReal.ofReal ((value P D .costs ξ').toReal - (value P D .costs ξ).toReal - ε) ≤
ENNReal.ofReal L * (inputDist P D .costs ξ ξ' +
∑ τ ∈ Finset.Icc 2 (D.T - 1), eLpNorm (xbar τ - P[xbar τ | sigmaGen ξ' τ]) ⊤ P) := by sorry
end MultistageStab.Value
Source
Heitsch, Römisch & Strugarek, Stability of multistage stochastic programs, author manuscript (edoc.hu-berlin.de, c. 2005), pp. 5–7, proof of Theorem 2.1, case "only random costs, i.e., r = 1 and r′ = ∞"
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.