Expected revenue equals
ProvedPersonalizedAssortment.IB.revenue_eq_expected_salesLet be any selection rule and let be the inventory process it generates. The expected revenue collected over the horizon is
It equals the revenue of the units sold by the end of the horizon:
This is the identity the proof of Theorem 1 uses for the numerator of the competitive ratio.
import Mathlib import Definitions.Def_PersonalizedAssortment_IB_ChoiceModel import Definitions.Def_PersonalizedAssortment_IB_InventoryBalancing
namespace PersonalizedAssortment.IB
/-- §A.3, p. 1549: the expected revenue of the algorithm equals `𝔼[∑_{i=1}^n r_i (c_i - I_i^T)]`,
the revenue of the units sold by the end of the horizon. -/
theorem revenue_eq_expected_sales {n : ℕ} {Z : Type*} (r : Fin n → ℝ) (c : Fin n → ℕ)
(φ : Z → ChoiceModel n) {T : ℕ} (z : Fin T → Z)
(sel : Fin T → (Fin n → ℕ) → Finset (Fin n)) :
ibRevenue r c φ z sel =
expectInv φ z sel c T le_rfl (fun I => ∑ i : Fin n, r i * ((c i : ℝ) - (I i : ℝ))) := by sorry
end PersonalizedAssortment.IB
Read-back
What the Lean code literally says, in plain math · claude-opus-5-5
Data.
- Products are and periods are .
- are arbitrary reals and are arbitrary naturals.
- assigns to each a choice model (nonnegative , zero for , with ).
- .
- is an arbitrary map from periods and vectors to subsets of .
No further hypotheses are made.
Expectation operator. is defined by the recursion
where and lowers coordinate by one, floored at . This is the expectation of for the chain started at that, in step from state , lowers coordinate with probability and otherwise stays.
Conclusion. The statement asserts
where is real subtraction.
Degenerate cases. If , both sides are . If , both sides are .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.