Proof of Theorem 2.1: is a convex risk measure generated by the convex, normalized penalty
ProvedSoftRobust.Duality.rhoQ_normalized_convexconvex-risk-measuresp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1penalty-function
Let be a nondecreasing, convex family of probability measures on with nonempty, and let . Then:
- , defined by , is a convex risk measure on bounded measurable payoffs (Definitions 2.1 and 2.2);
- is generated by , and is convex in the sense of mixtures:
- is normalized: for some .
This is the part of Theorem 2.1 that identifies as "the normalized, convex risk measure generated by the penalty function ".
Formalization Note takes values in , so nonnegativity is automatic. "Generated by" includes the least-upper-bound property of the supremum for every bounded measurable .
Preamble
import Mathlib import Definitions.Def_SoftRobust_Duality_Mixture import Definitions.Def_SoftRobust_Duality_ConvexRiskMeasure import Definitions.Def_SoftRobust_Duality_Model open MeasureTheory open scoped NNReal ENNReal
Formal statement
namespace SoftRobust.Duality
/-- §2.3, proof of Theorem 2.1, p. 1224: for a nondecreasing, convex family `𝐐` with `𝒬(0)`
nonempty, `ρ^Q` is a convex risk measure generated by the convex penalty `α^Q`, and `α^Q` is
normalized. -/
theorem rhoQ_normalized_convex {Ω : Type*} [MeasurableSpace Ω]
(𝒬 : ℝ≥0 → Set (ProbabilityMeasure Ω)) (h𝒬 : IsAdmissibleFamily 𝒬) :
IsConvexRiskMeasure (rhoQ 𝒬) ∧ IsGeneratedBy (rhoQ 𝒬) (penaltyOfFamily 𝒬) ∧
IsNormalizedPenalty (penaltyOfFamily 𝒬) := by sorry
end SoftRobust.Duality
Source
Ben-Tal, Bertsimas, Brown, A Soft Robust Model for Optimization Under Ambiguity, Oper. Res. 58(4) (2010), p. 1224, §2.3, proof of Theorem 2.1 (convexity of α^Q and ρ^Q, normalization)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.