Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 7, p. 21 — scaling the nested-by-revenue LP optimum by the factor α of (6) gives a feasible solution of (3)

Proved
NestedLogitVariants.Synergistic.theorem_7

by mikedeng1 · Oct 5, 2026 · Mathlib 0df444a (Lean v4.33.1)

assortment-optimizationnested-by-revenuenested-logitp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1performance-guaranteerevenue-management

Consider the nested logit assortment problem with fully-captured nests (vi0=0v_{i0} = 0vi0​=0 for every nest iii), γi>1\gamma_i > 1γi​>1 for some nest iii, positive revenues and n≥2n \ge 2n≥2 products per nest. Let (x^,y^)(\hat x, \hat y)(x^,y^​) be an optimal solution of the linear program (4) in which the candidate assortments of every nest are the nested-by-revenue assortments {Nij:j∈N+}\{N_{ij} : j \in N_+\}{Nij​:j∈N+​}, and let

α=max⁡i∈M, j=2,…,n{Ri(Ni,j−1)Ri(Nij)∧(Ri(Nij)Ri(Ni,j−1) Vi(Nij)γiVi(Ni,j−1)γi)}(6)\alpha = \max_{i \in M,\ j = 2, \dots, n} \left\{\frac{R_i(N_{i,j-1})}{R_i(N_{ij})} \wedge \left(\frac{R_i(N_{ij})}{R_i(N_{i,j-1})}\, \frac{V_i(N_{ij})^{\gamma_i}}{V_i(N_{i,j-1})^{\gamma_i}}\right)\right\} \tag{6}α=i∈M, j=2,…,nmax​{Ri​(Nij​)Ri​(Ni,j−1​)​∧(Ri​(Ni,j−1​)Ri​(Nij​)​Vi​(Ni,j−1​)γi​Vi​(Nij​)γi​​)}(6)

with a∧b=min⁡{a,b}a \wedge b = \min\{a, b\}a∧b=min{a,b}. Then (αx^,αy^)(\alpha \hat x, \alpha \hat y)(αx^,αy^​) is feasible for problem (3):

v0 αx^≥∑i∈Mαy^i,αy^i≥Vi(Si)γi (Ri(Si)−αx^)∀Si⊆N, i∈M.v_0\, \alpha \hat x \ge \sum_{i \in M} \alpha \hat y_i, \qquad \alpha \hat y_i \ge V_i(S_i)^{\gamma_i}\,(R_i(S_i) - \alpha \hat x) \quad \forall S_i \subseteq N,\ i \in M.v0​αx^≥i∈M∑​αy^​i​,αy^​i​≥Vi​(Si​)γi​(Ri​(Si​)−αx^)∀Si​⊆N, i∈M.

Combined with Theorem 1, this shows that the best combination of nested-by-revenue assortments, found by a linear program with 1+m1 + m1+m variables and 1+m(1+n)1 + m(1 + n)1+m(1+n) constraints, earns at least a fraction 1/α1/\alpha1/α of the optimal expected revenue.

Formalization Note The standing assumptions are those of the model (see the Model definition) and of §4: every nest is fully captured, vi0=0v_{i0} = 0vi0​=0, and γi>1\gamma_i > 1γi​>1 for some nest iii (p. 16). The positive revenues rij>0r_{ij} > 0rij​>0 are the disclosed pin of this mission: they keep every denominator Ri(Nij)R_i(N_{ij})Ri​(Nij​), j≥1j \ge 1j≥1, of (6) positive. The condition n≥2n \ge 2n≥2 makes the range of (6) nonempty. The number α\alphaα is given together with the hypothesis that it is the greatest element of the set of terms of (6), not merely an upper bound. The collection {Nij:j∈N+}\{N_{ij} : j \in N_+\}{Nij​:j∈N+​} is the set of nbr n j for j≤nj \le nj≤n, including Ni0=∅N_{i0} = \emptysetNi0​=∅. An optimal solution of (4) is a feasible pair whose xxx is no larger than that of any feasible pair.


Retired 2026-10-27: a Lean-vs-paper fidelity audit found this statement weaker than, or narrower than, the cited result. A corrected statement will be resubmitted.

Preamble
import Mathlib
import Definitions.Def_NestedLogitVariants_Synergistic_Factor
Formal statement
namespace NestedLogitVariants.Synergistic

/-- Theorem 7, p. 21: let `(x̂, ŷ)` be an optimal solution of problem (4) with the candidate
collections replaced by the nested-by-revenue assortments `{N_{ij} : j ∈ N_+}`. Then, with `α` the
expression in (6), `(α x̂, α ŷ)` is a feasible solution of problem (3). -/
theorem theorem_7 {ι : Type*} [Fintype ι] [DecidableEq ι] {n : ℕ} (I : Instance ι n)
    (hI : I.Standing) (hfc : ∀ i, I.vnp i = 0) (hsyn : ∃ i, 1 < I.γ i)
    (hr : ∀ i j, 0 < I.r i j) (hn : 2 ≤ n)
    (α : ℝ) (hα : IsGreatest (alphaTerms I) α)
    (xh : ℝ) (yh : ι → ℝ) (hopt : LP4Optimal I (fun _ => {S | ∃ j ≤ n, S = nbr n j}) xh yh) :
    LP3Feasible I (α * xh) (α • yh) := by sorry

end NestedLogitVariants.Synergistic
Source
Davis, Gallego, Topaloglu, Assortment optimization under variants of the nested logit model, revised manuscript of June 18, 2013 (published Oper. Res. 62(2), 2014, DOI 10.1287/opre.2014.1256), p. 21, Theorem 7 (proof in Appendix A.1, pp. 41–43)
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