§3.1, p. 588 — the dual matrix Y of a feasible point of (12) is positive semidefinite
ProvedMinimaxSLP.RhsSDP.dual_Y_posSemidefdualityp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1semidefinite-programming
Let for all , and let satisfy complete recourse (Assumption 2) and (Assumption 3 at the constant ). If is feasible for the dual (12) at some , that is, is symmetric and for all , then
This is the observation that the dual feasible set only contains positive semidefinite matrices, used before the separation problem (13).
Preamble
import Mathlib import Definitions.Def_MomentDRO_Conf_Setting import Definitions.Def_WorstCaseVaR_KnownMoments_Basic import Definitions.Def_MinimaxSLP_RhsSDP_Model
Formal statement
namespace MinimaxSLP.RhsSDP
open MeasureTheory Matrix
/-- §3.1, p. 588: the dual matrix `Y` of every feasible point `(Y, y, y₀)` of (12) is positive
semidefinite. Stated under `α_k ≥ 0`, complete recourse (Assumption 2) and
`{π : W′π ≤ q} ≠ ∅` (Assumption 3 at the constant `q`). -/
theorem dual_Y_posSemidef {r d n K : ℕ} [NeZero K]
(W : Matrix (Fin r) (Fin d) ℝ) (T : Matrix (Fin r) (Fin n) ℝ) (q : Fin d → ℝ)
(α β : Fin K → ℝ)
(hα : ∀ k, 0 ≤ α k)
(hA2 : ∀ z : Fin r → ℝ, ∃ w : Fin d → ℝ, 0 ≤ w ∧ W *ᵥ w = z)
(hA3 : ∃ π : Fin r → ℝ, Wᵀ *ᵥ π ≤ q)
(x : Fin n → ℝ) (s : Matrix (Fin r) (Fin r) ℝ × (Fin r → ℝ) × ℝ)
(hs : s ∈ dualFeasible12 W T q α β x) :
s.1.PosSemidef := by sorry
end MinimaxSLP.RhsSDP
Source
Bertsimas, Doan, Natarajan & Teo, Models for Minimax Stochastic Linear Optimization Problems with Risk Aversion, Math. Oper. Res. 35(3), 2010, p. 588, §3.1
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.