Proof of Theorem 2.1, converse: is nondecreasing and convex, and
ProvedSoftRobust.Duality.familyOfPenalty_propsconvex-risk-measuresp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1penalty-function
Let be a penalty on the probability measures of that is convex in the sense of mixtures and normalized ( for some ). Define
Then the family is nondecreasing and convex (if , then for ), and is nonempty.
This is the first step of the converse half of Theorem 2.1: every normalized convex penalty produces a family to which the forward half applies.
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 (converse, first step): for a convex, normalized penalty
`α`, the family `𝐐^α = {𝒬^α(ε) : ε ≥ 0}`, `𝒬^α(ε) = {ℚ : α(ℚ) ≤ ε}`, is nondecreasing and convex, and
`𝒬^α(0)` is nonempty. -/
theorem familyOfPenalty_props {Ω : Type*} [MeasurableSpace Ω]
(α : ProbabilityMeasure Ω → ℝ≥0∞) (hconv : IsConvexPenalty α)
(hnorm : IsNormalizedPenalty α) :
IsAdmissibleFamily (familyOfPenalty α) := 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, converse
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.