Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Sec. 2.2.2, p. 9 — the hindsight optimum is at most the DLP value: vHO≤vDLPv^{\mathrm{HO}} \le v^{\mathrm{DLP}}vHO≤vDLP

Proved
ResolvingNRM.IRT.hindsight_le_dlp

by mikedeng1 · Oct 6, 2026 · Mathlib 0df444a (Lean v4.33.1)

linear-programmingnetwork-revenue-managementp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1

Consider the network revenue-management model with nnn classes of Poisson rates λj>0\lambda_j > 0λj​>0, revenues rj≥0r_j \ge 0rj​≥0 and a nonnegative bill-of-materials matrix A∈Rm×nA \in \mathbb R^{m\times n}A∈Rm×n. For every horizon T>0T > 0T>0 and every capacity vector C≥0C \ge 0C≥0, the hindsight optimum is bounded by the value of the deterministic LP:

vHO(T,C)=E[VHO]  ≤  vDLP(T,C)=Tmax⁡{∑jrjxj ∣ ∑jAjxj≤C/T, 0≤xj≤λj}.v^{\mathrm{HO}}(T, C) = \mathbb E\big[V^{\mathrm{HO}}\big] \;\le\; v^{\mathrm{DLP}}(T, C) = T\max\Big\{\sum_j r_j x_j \ \Big|\ \sum_j A_j x_j \le C/T,\ 0 \le x_j \le \lambda_j\Big\}.vHO(T,C)=E[VHO]≤vDLP(T,C)=Tmax{j∑​rj​xj​ ​ j∑​Aj​xj​≤C/T, 0≤xj​≤λj​}.

Together with v∗≤vHOv^* \le v^{\mathrm{HO}}v∗≤vHO this makes the hindsight optimum a tighter upper bound than the DLP; it is what turns Proposition 5's bound against the DLP into a bound on the regret against the hindsight optimum.

Formalization Note Only the second inequality of the display v∗≤vHO≤vDLPv^* \le v^{\mathrm{HO}} \le v^{\mathrm{DLP}}v∗≤vHO≤vDLP is stated: the optimal value v∗v^*v∗ over all admissible policies is not defined in this mission. The positivity of the rates and the nonnegativity of rrr and AAA are the standing assumptions of Sec. 2, p. 7, which the page leaves implicit. The horizon TTT is real.

Preamble
import Mathlib
import Definitions.Def_RLPBidPrice_Unbiased_Model
import Definitions.Def_ResolvingNRM_IRT_Model

open RLPBidPrice.Unbiased Matrix
Formal statement
namespace ResolvingNRM.IRT

theorem hindsight_le_dlp {m n : ℕ} (A : Matrix (Fin m) (Fin n) ℝ) (r lam : Fin n → ℝ)
    (hlam : ∀ j, 0 < lam j) (hr : ∀ j, 0 ≤ r j) (hA : ∀ l j, 0 ≤ A l j)
    (T : ℝ) (hT : 0 < T) (C : Fin m → ℝ) (hC : 0 ≤ C) :
    hindsightValue A r lam T C ≤ dlpValue A r lam T C := by sorry

end ResolvingNRM.IRT
Source
Bumpensanti, Wang, A Re-solving Heuristic with Uniformly Bounded Loss for Network Revenue Management, arXiv:1802.06192v3, Sec. 2.2.2, p. 9, display 'v* ≤ v^HO ≤ v^DLP'
Human review
  • Endorsed by Shuze Chen · Oct 6, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 6, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me