Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 1, p. 7 — SampleGreedy is a (p+1)²/p-approximation (p+1 monotone, q = 1/(p+1); p linear, q = 1/p) over p-extendible systems

Proved
SubsampleSubmod.SampleGreedy.theorem_1

by mikedeng1 · Oct 22, 2026 · Mathlib 0df444a (Lean v4.33.1)

approximation-algorithmsp-extendible-systemsp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1randomized-algorithmssubmodular-maximization

Let N\mathcal NN be a finite ground set with a fixed linear order (used to break ties), let f:2N→R≥0f : 2^{\mathcal N} \to \mathbb R_{\ge 0}f:2N→R≥0​ be a non-negative submodular function, and let (N,I)(\mathcal N, \mathcal I)(N,I) be a ppp-extendible independence system with p≥1p \ge 1p≥1. Write Eq[f(S)]\mathbb E_q[f(S)]Eq​[f(S)] for the expected value of the set SSS returned by SampleGreedy (Algorithm 1) when every element is sampled independently with probability qqq. Then:

  1. SampleGreedy always returns an independent set: S∈IS \in \mathcal IS∈I for every sample.
  2. With q=1p+1q = \frac{1}{p+1}q=p+11​, for every independent set T∈IT \in \mathcal IT∈I,
f(T)  ≤  (p+1)2p  E1/(p+1)[f(S)].f(T) \;\le\; \frac{(p+1)^2}{p}\; \mathbb E_{1/(p+1)}[f(S)].f(T)≤p(p+1)2​E1/(p+1)​[f(S)].
  1. If moreover fff is monotone, then with q=1p+1q = \frac{1}{p+1}q=p+11​, f(T)≤(p+1) E1/(p+1)[f(S)]f(T) \le (p+1)\, \mathbb E_{1/(p+1)}[f(S)]f(T)≤(p+1)E1/(p+1)​[f(S)] for every T∈IT \in \mathcal IT∈I.
  2. If moreover fff is linear, then with q=1pq = \frac1pq=p1​, f(T)≤p E1/p[f(S)]f(T) \le p\, \mathbb E_{1/p}[f(S)]f(T)≤pE1/p​[f(S)] for every T∈IT \in \mathcal IT∈I.

Taking TTT to be a maximizer OPT of problem (2), these say that SampleGreedy is, in expectation, a (p+1)2/p(p+1)^2/p(p+1)2/p-approximation for non-negative submodular maximization over ppp-extendible systems, a (p+1)(p+1)(p+1)-approximation for monotone objectives, and a ppp-approximation for linear objectives. Since no polynomial-query algorithm beats p+1/2p + 1/2p+1/2 for this problem, the ratio (p+1)2/p=p+2+1/p(p+1)^2/p = p + 2 + 1/p(p+1)2/p=p+2+1/p is nearly tight.

Formalization Note The approximation ratio of a randomized algorithm is read as the paper's proof reads it: E[f(S)]≥f(OPT)/α\mathbb E[f(S)] \ge f(\mathrm{OPT})/\alphaE[f(S)]≥f(OPT)/α. Comparing with every T∈IT \in \mathcal IT∈I is equivalent to comparing with OPT. Additions relative to the page: ppp is a natural number with p≥1p \ge 1p≥1 (the paper divides by ppp); ties are broken by a fixed linear order of the ground set, as the paper assumes on p. 10; the expectation is the exact sum over all 2n2^n2n samples; the while loop runs nnn iterations. "Linear" is the page's definition ((1) with equality) and does not force f(∅)=0f(\emptyset) = 0f(∅)=0. Submodularity is the published lattice form, equivalent on finite ground sets to the diminishing-returns form (1). The oracle-count clause "SampleGreedy uses in expectation O(nk/p)O(nk/p)O(nk/p) calls to the evaluation and independence oracles" is not formalized.

Preamble
import Mathlib
import Definitions.Def_SubsampleSubmod_SampleGreedy_Model
Formal statement
namespace SubsampleSubmod.SampleGreedy

/-- Theorem 1 (Harshaw et al., arXiv:2104.02772v1, p. 7). Let `f ≥ 0` be submodular and `ℐ` a
`p`-extendible independence system, `p ≥ 1`. SampleGreedy always returns an independent set, and
1. with `q = 1/(p+1)`: `f(T) ≤ (p+1)²/p · E[f(S)]` for every `T ∈ ℐ`;
2. with `q = 1/(p+1)` and `f` monotone: `f(T) ≤ (p+1) · E[f(S)]` for every `T ∈ ℐ`;
3. with `q = 1/p` and `f` linear: `f(T) ≤ p · E[f(S)]` for every `T ∈ ℐ`.
The oracle-count clause `O(nk/p)` is not formalized. -/
theorem theorem_1 {α : Type} [Fintype α] [LinearOrder α] (f : Finset α → ℝ)
    (I : Finset α → Prop) (p : ℕ) (hp : 1 ≤ p) (hf0 : ∀ S, 0 ≤ f S)
    (hsub : NonmonotoneSubmod.Shared.Submodular f) (hI : IsIndepSystem I)
    (hext : IsPExtendible p I) :
    (∀ N' : Finset α, I (sampleGreedy f I N')) ∧
    (∀ T : Finset α, I T →
      f T ≤ (((p : ℝ) + 1) ^ 2 / (p : ℝ)) * expectedValue (1 / ((p : ℝ) + 1)) f I) ∧
    (Monotone f → ∀ T : Finset α, I T →
      f T ≤ ((p : ℝ) + 1) * expectedValue (1 / ((p : ℝ) + 1)) f I) ∧
    (IsLinear f → ∀ T : Finset α, I T →
      f T ≤ (p : ℝ) * expectedValue (1 / (p : ℝ)) f I) := by sorry

end SubsampleSubmod.SampleGreedy
Source
Harshaw et al., arXiv:2104.02772v1, p. 7, Theorem 1 (proof §4.1, pp. 9–17: §4.1.1 pp. 14–15 submodular, §4.1.2 p. 17 linear)
Human review
  • Endorsed by Shuze Chen · Oct 23, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 23, 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