(24), Sect. 5.1.4, p. 18 — inf over feasible polynomials of p(1) equals γ*
ProvedWaitJudge.Convex.inf_poly_eq_gammaStarcalculus-of-variationsmoment-problemp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1polynomial-approximation
Let be natural numbers and . Let be the value of the variational problem (10): the infimum of over the functions with
Then the infimum of over the polynomials satisfying the same constraints equals :
Here is the value of the dual problem (21) rewritten over polynomials of degree , and the union over of these classes is the set of all real polynomials. The equality identifies the bound obtained by duality with the value of (10).
Formalization Note is ContDiffOn ℝ d on with derivatives taken within ; a polynomial is feasible when its evaluation map is. Both infima are real sInfs of nonempty sets bounded below by .
Preamble
import Mathlib import Definitions.Def_WaitJudge_Convex_Setting
Formal statement
namespace WaitJudge.Convex
theorem inf_poly_eq_gammaStar (N d : ℕ) (ε : ℕ → ℝ) (hε : ∀ k ≤ d, 0 ≤ ε k ∧ ε k ≤ 1) :
sInf {y | ∃ q : Polynomial ℝ, IsFeasible10 N d ε (fun t => q.eval t) ∧ y = q.eval 1} =
gammaStar N d ε := by sorry
end WaitJudge.Convex
Source
Campi & Garatti, Wait-and-judge scenario optimization, Math. Program. (2018), doi:10.1007/s10107-016-1056-9 (accepted manuscript), PDF pp. 17–18, Sect. 5.1.4, rewriting of (21) over P_M and (24)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.