Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 3, p. 18 — under Assumption 1, εi∗=yi∗\varepsilon^*_i = y^*_iεi∗​=yi∗​ and weights yi∗/∑jyj∗y^*_i/\sum_j y^*_jyi∗​/∑j​yj∗​ (y∗y^*y∗ solving (23)) give the optimal central mechanism

Proved
PrivAcq.Central.theorem_3

by mikedeng1 · Oct 21, 2026 · Mathlib 0df444a (Lean v4.33.1)

data-acquisitiondifferential-privacymechanism-designoptimal-mechanismp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1

Consider the central-privacy data acquisition model under the standing assumptions, and suppose Assumption 1 holds: every virtual cost ψi(c)=c+Fi(c)/fi(c)\psi_i(c) = c + F_i(c)/f_i(c)ψi​(c)=c+Fi​(c)/fi​(c) is increasing. For every type profile ccc let y∗(c)=(y1∗(c),…,yn∗(c))y^*(c) = (y^*_1(c),\dots,y^*_n(c))y∗(c)=(y1∗​(c),…,yn∗​(c)) be an optimal solution of

min⁡y  n+1(∑j=1nyj)2(2+∑i=1nyi2 VAR)+∑i=1nψi(ci) yis.t. yi≥0 for all i,(23)\min_{y}\; \frac{n+1}{\big(\sum_{j=1}^n y_j\big)^2}\Big(2 + \sum_{i=1}^n y_i^2\,\mathrm{VAR}\Big) + \sum_{i=1}^n \psi_i(c_i)\,y_i \quad \text{s.t. } y_i \ge 0 \text{ for all } i, \qquad (23)ymin​(∑j=1n​yj​)2n+1​(2+i=1∑n​yi2​VAR)+i=1∑n​ψi​(ci​)yi​s.t. yi​≥0 for all i,(23)

chosen measurably in ccc. Let the mechanism M∗M^*M∗ use privacy losses εi∗(c)=yi∗(c)\varepsilon^*_i(c) = y^*_i(c)εi∗​(c)=yi∗​(c), the estimator weights wi∗(c)=yi∗(c)/∑j=1nyj∗(c)w^*_i(c) = y^*_i(c)/\sum_{j=1}^n y^*_j(c)wi∗​(c)=yi∗​(c)/∑j=1n​yj∗​(c) and Laplace parameter η∗(c)=∑jyj∗(c)\eta^*(c) = \sum_j y^*_j(c)η∗(c)=∑j​yj∗​(c), and the payments (19) with di=0d_i = 0di​=0. Then

  1. M∗M^*M∗ is an admissible mechanism of class (22), incentive compatible (17) and individually rational (18); and
  2. its expected platform cost Ec[MSE(c)+∑iti(c)]\mathbb E_c[\mathrm{MSE}(c) + \sum_i t_i(c)]Ec​[MSE(c)+∑i​ti​(c)] is at most that of every admissible mechanism of class (22) that is incentive compatible and individually rational.

So the optimal privacy loss level is εi∗(c)=yi∗\varepsilon^*_i(c) = y^*_iεi∗​(c)=yi∗​, and user iii's data receives weight yi∗/∑jyj∗y^*_i/\sum_j y^*_jyi∗​/∑j​yj∗​ in the platform's estimator.

Formalization Note The theorem is proved for every measurable selection of minimizers (the minimizer need not be unique); it does not claim that every optimal mechanism coincides with M∗M^*M∗. The feasible set of (23) adds ∑jyj>0\sum_j y_j > 0∑j​yj​>0 (at y=0y = 0y=0 the objective is undefined). Model conventions (compact type support with ai>0a_i > 0ai​>0, no differentiability or sign constraint on payments, strictly increasing ψi\psi_iψi​) are listed in PrivAcq.Central.Model.

Preamble
import Mathlib
import Definitions.Def_PrivAcq_Central_Model
Formal statement
namespace PrivAcq.Central

open MeasureTheory

theorem theorem_3 {n : ℕ} (S : Setting n) (hS : S.Standing) (hA1 : S.Assumption1)
    (ystar : (Fin n → ℝ) → Fin n → ℝ) (hmeas : Measurable ystar)
    (hopt : ∀ c ∈ S.types, IsOpt23 S c (ystar c)) :
    Admissible S (optMech S ystar) ∧ IC S (optMech S ystar) ∧ IR S (optMech S ystar) ∧
      ∀ M : Mech n, Admissible S M → IC S M → IR S M →
        objective S (optMech S ystar) ≤ objective S M := by sorry

end PrivAcq.Central
Source
Fallah, Makhdoumi, Malekian, Ozdaglar, arXiv:2201.03968v2, p. 18, Theorem 3 and (23); proof pp. 44–45
Human review
  • Endorsed by Shuze Chen · Oct 21, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 21, 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