Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proposition 3.1 — the Laplace transform method: P{λmax(Y) ≥ t} ≤ e^{−θt}·E tr e^{θY}

Proved
MatrixTail.Master.proposition3_1

by mikedeng1 · Oct 7, 2026 · Mathlib 0df444a (Lean v4.33.1)

laplace-transformp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1random-matricestail-bound

Let YYY be a random d×dd\times dd×d complex Hermitian matrix on a probability space, with d≥1d\ge 1d≥1. For every t∈Rt\in\mathbb Rt∈R and every θ>0\theta>0θ>0 at which Etr⁡eθY\mathbb E \operatorname{tr} e^{\theta Y}EtreθY exists,

P{λmax⁡(Y)≥t}≤e−θt⋅Etr⁡eθY.\mathbb P\{\lambda_{\max}(Y)\ge t\} \le e^{-\theta t}\cdot \mathbb E\operatorname{tr} e^{\theta Y}.P{λmax​(Y)≥t}≤e−θt⋅EtreθY.

Equivalently, P{λmax⁡(Y)≥t}≤inf⁡θ>0{e−θt Etr⁡eθY}\mathbb P\{\lambda_{\max}(Y)\ge t\}\le \inf_{\theta>0}\{e^{-\theta t}\,\mathbb E\operatorname{tr}e^{\theta Y}\}P{λmax​(Y)≥t}≤infθ>0​{e−θtEtreθY}, the form in which the paper states it.

This is the matrix analogue of the scalar Laplace transform (Chernoff) bound: a tail bound for the largest eigenvalue reduces to a bound on the trace of the matrix moment generating function.

Formalization Note. The infimum is stated as the bound for each θ>0\theta>0θ>0 at which ω↦tr⁡eθY(ω)\omega\mapsto\operatorname{tr}e^{\theta Y(\omega)}ω↦treθY(ω) is integrable; the paper admits expectations that do not exist (value +∞+\infty+∞), so this reading is exact. The dimension is assumed positive: for d=0d=0d=0 the largest eigenvalue would be 000 and the trace 000.

Preamble
import Mathlib
import Definitions.Def_MatrixTail_Master_Model

open MeasureTheory ProbabilityTheory
open scoped MatrixOrder ComplexOrder ENNReal
Formal statement
namespace MatrixTail.Master

/-- **Proposition 3.1 (The Laplace Transform Method).** Tropp, *User-Friendly Tail Bounds for Sums of Random
Matrices*, arXiv:1004.4389v7, p. 9: "Let `Y` be a random self-adjoint matrix. For all `t ∈ ℝ`,
`P {λmax(Y) ≥ t} ≤ inf_{θ>0} { e^{−θt} · E tr e^{θY} }`."

**Formalization Note.**
* Matrices are complex `d × d` with `d ≥ 1` (`[NeZero d]`); for `d = 0`, `λmax = sSup ∅ = 0` and the trace
  is `0`, and the bound would fail at `t ≤ 0`.
* `Y` is measurable and Hermitian at every outcome; `P` is a probability measure.
* The infimum over `θ > 0` is stated as the bound for every `θ > 0` at which `E tr e^{θY}` exists, i.e. at
  which `ω ↦ tr e^{θY(ω)}` is integrable (the §2.2 regularity made explicit). The paper admits non-existent
  expectations, whose value is then `+∞`; `P ≤ inf_θ F(θ)` is equivalent to `P ≤ F(θ)` for every `θ`, so this
  reading is exact. Without the integrability hypothesis Lean's integral would be `0`.
* `e^{θY}` is `cfc Real.exp (θ • Y)` and `tr e^{θY}` is the real part of its trace (`trExp`). -/
theorem proposition3_1 {Ω : Type*} [MeasurableSpace Ω] (P : Measure Ω) [IsProbabilityMeasure P]
    {d : ℕ} [NeZero d] (Y : Ω → Matrix (Fin d) (Fin d) ℂ)
    (hY_meas : Measurable Y) (hY_herm : ∀ ω, (Y ω).IsHermitian)
    (t θ : ℝ) (hθ : 0 < θ) (hint : Integrable (fun ω => trExp (θ • Y ω)) P) :
    P.real {ω | t ≤ lambdaMax (Y ω)} ≤ Real.exp (-θ * t) * ∫ ω, trExp (θ • Y ω) ∂P := by sorry

end MatrixTail.Master
Source
Tropp, User-Friendly Tail Bounds for Sums of Random Matrices, arXiv:1004.4389v7, p. 9, Proposition 3.1
Human review
  • Endorsed by Shuze Chen · Oct 8, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 8, 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