Proof of Theorem 3.8, p. 12 — Y*(V) = max{λ : max_{|S|≤K} min_{v∈V} (1/v₀) Σ_{i∈S} v_i (r_i − λ) ≥ λ}
ProvedRobustMNL.SizeCon.parametric_formLet be a compact, nonempty set of positive MNL parameter vectors , let be real revenues and a size limit. Write for the optimal value of the size-constrained robust problem. Then is the largest real number satisfying
that is,
This parametric (Dinkelbach-type) description of holds for every compact uncertainty set; it is the first step of the proof of Theorem 3.8, which then evaluates the left-hand side in closed form when is a box.
Formalization Note The "max" of the set of admissible is IsGreatest: belongs to the set and bounds every member, so the maximum is attained. The inner is a real infimum over , attained under the stated hypotheses (compact, nonempty, every component positive). Revenues are arbitrary reals; the paper's ordering is not assumed. Products are 0-indexed (Fin n).
import Mathlib import Definitions.Def_ChoiceCDLP_MNL_mnlObjective import Definitions.Def_RobustMNL_SizeCon_Model
namespace RobustMNL.SizeCon
/-- Proof of Theorem 3.8, p. 12, first display: for every compact, nonempty uncertainty set
`V ⊆ ℝⁿ⁺¹₊₊`, `Y*(V)` is the largest `λ` with
`λ ≤ max_{S : |S| ≤ K} min_{v ∈ V} (1/v₀) ∑_{i ∈ S} v_i (r_i − λ)`. -/
theorem parametric_form {n : ℕ} (V : Set (ℝ × (Fin n → ℝ))) (hVc : IsCompact V)
(hVne : V.Nonempty) (hVpos : ∀ p ∈ V, RobustMNL.Static.IsPos p) (r : Fin n → ℝ) (K : ℕ) :
IsGreatest {lam : ℝ | lam ≤ paramValue V r K lam} (Ystar V r K) := by sorry
end RobustMNL.SizeCon
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.