Proof of Lemma 3.1, p. 6 — f(A ∪ {i}, v) is a convex combination of rᵢ and f(A, v)
ProvedRobustMNL.Static.convex_combinationassortment-optimizationmultinomial-logitp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1
Let , let be an assortment and let be a product. Then the expected revenue of is the following convex combination of the revenue and the expected revenue :
The two weights are positive and add up to one, so lies between and . This identity is the whole content of Lemma 3.1.
Formalization Note Products are Fin n (Lean index is the paper's product ); is p.1 and is p.2 i. The revenues are arbitrary reals, a harmless generalization of the paper's .
Preamble
import Mathlib import Definitions.Def_RobustMNL_Static_Model
Formal statement
namespace RobustMNL.Static
theorem convex_combination {n : ℕ} (r : Fin n → ℝ) (p : ℝ × (Fin n → ℝ)) (hp : IsPos p)
(A : Finset (Fin n)) (i : Fin n) (hi : i ∉ A) :
rev r (insert i A) p =
p.2 i / (p.1 + p.2 i + ∑ ℓ ∈ A, p.2 ℓ) * r i
+ (p.1 + ∑ ℓ ∈ A, p.2 ℓ) / (p.1 + p.2 i + ∑ ℓ ∈ A, p.2 ℓ) * rev r A p := by sorry
end RobustMNL.Static
Source
Rusmevichientong, Topaloglu, Robust Assortment Optimization in Revenue Management Under the Multinomial Logit Choice Model, Operations Research (2012), doi:10.1287/opre.1120.1063, authors' manuscript of 20 Sep 2011, Proof of Lemma 3.1, p. 6, display
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.