Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 10.5 — concentration of the random assignment cost L_N around its median

Proved
TalagrandConc.Assignment.theorem_10_5

by mikedeng1 · Oct 6, 2026 · Mathlib 0df444a (Lean v4.33.1)

assignment-problemconcentration-of-measurep2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-bookp2o-v1probability

Let the costs Xi,jX_{i,j}Xi,j​, i∈Ii \in Ii∈I, j∈Jj \in Jj∈J, ∣I∣=∣J∣=N|I| = |J| = N∣I∣=∣J∣=N, be independent and uniformly distributed on [0,1][0,1][0,1], and let

LN=min⁡{∑i∈IXi,τ(i)  ;  τ:I→J one-to-one}L_N = \min\Big\{ \sum_{i \in I} X_{i,\tau(i)} \;;\; \tau : I \to J \text{ one-to-one} \Big\}LN​=min{i∈I∑​Xi,τ(i)​;τ:I→J one-to-one}

be the optimal assignment cost. There is a universal constant K>0K > 0K>0 such that for every N≥3N \ge 3N≥3, every median MMM of LNL_NLN​ and every t≥0t \ge 0t≥0:

t≤log⁡N⇒P(∣LN−M∣≥Kt(log⁡N)2Nlog⁡log⁡N)≤2exp⁡(−t2),(10.7)t \le \sqrt{\log N} \Rightarrow P\Big(|L_N - M| \ge \frac{K t (\log N)^2}{\sqrt N \log\log N}\Big) \le 2\exp(-t^2), \tag{10.7}t≤logN​⇒P(∣LN​−M∣≥N​loglogNKt(logN)2​)≤2exp(−t2),(10.7) t≥log⁡N⇒P(∣LN−M∣≥Kt3log⁡NNlog⁡t2)≤2exp⁡(−t2).(10.8)t \ge \sqrt{\log N} \Rightarrow P\Big(|L_N - M| \ge \frac{K t^3 \log N}{\sqrt N \log t^2}\Big) \le 2\exp(-t^2). \tag{10.8}t≥logN​⇒P(∣LN​−M∣≥N​logt2Kt3logN​)≤2exp(−t2).(10.8)

Since E(LN)E(L_N)E(LN​) stays bounded as N→∞N \to \inftyN→∞, this shows that the fluctuations of LNL_NLN​ are at most of order (log⁡N)2/(Nlog⁡log⁡N)(\log N)^2/(\sqrt N \log\log N)(logN)2/(N​loglogN), much smaller than its mean.

Formalization Note One constant KKK serves both (10.7) and (10.8) and is chosen before NNN, MMM and ttt. A median is any MMM with P(LN≤M)≥1/2P(L_N \le M) \ge 1/2P(LN​≤M)≥1/2 and P(LN≥M)≥1/2P(L_N \ge M) \ge 1/2P(LN​≥M)≥1/2. The paper leaves t≥0t \ge 0t≥0 implicit; for negative ttt (10.7) would fail. For N≥3N \ge 3N≥3, log⁡log⁡N>0\log\log N > 0loglogN>0, and t≥log⁡N>1t \ge \sqrt{\log N} > 1t≥logN​>1 gives log⁡t2>0\log t^2 > 0logt2>0.

Preamble
import Mathlib
import Definitions.Def_TalagrandConc_Assignment_Basic
Formal statement
namespace TalagrandConc.Assignment

open MeasureTheory

/-- Talagrand (1995), p. 169, Theorem 10.5. There is a universal constant `K > 0` such that
for every `N ≥ 3`, every median `M` of the optimal assignment cost `L_N` (costs i.i.d.
uniform on `[0, 1]`) and every `t ≥ 0`:
(10.7) if `t ≤ √(log N)` then `P(|L_N − M| ≥ K t (log N)² / (√N log log N)) ≤ 2 exp(−t²)`;
(10.8) if `t ≥ √(log N)` then `P(|L_N − M| ≥ K t³ log N / (√N log t²)) ≤ 2 exp(−t²)`. -/
theorem theorem_10_5 :
    ∃ K : ℝ, 0 < K ∧ ∀ N : ℕ, 3 ≤ N → ∀ M : ℝ, TalagrandConc.BinPacking.IsMedian (costLaw N) optCost M →
      ∀ t : ℝ, 0 ≤ t →
        (t ≤ Real.sqrt (Real.log N) →
          costLaw N {X | K * t * Real.log N ^ 2 / (Real.sqrt N * Real.log (Real.log N)) ≤
              |optCost X - M|} ≤ ENNReal.ofReal (2 * Real.exp (-t ^ 2))) ∧
        (Real.sqrt (Real.log N) ≤ t →
          costLaw N {X | K * t ^ 3 * Real.log N / (Real.sqrt N * Real.log (t ^ 2)) ≤
              |optCost X - M|} ≤ ENNReal.ofReal (2 * Real.exp (-t ^ 2))) := by sorry

end TalagrandConc.Assignment
Source
Talagrand, Concentration of measure and isoperimetric inequalities in product spaces, Publ. Math. IHÉS 81 (1995), p. 169, Theorem 10.5, Eqs. (10.7)–(10.8)
Human review
  • Endorsed by Shuze Chen · Oct 6, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 6, 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