Theorem 6.6.1 — Hermitian matrix Bernstein
ProvedTroppMatrixConcentration.hermitian_bernsteinLet be a finite independent family of measurable Hermitian random matrices with finite second moments, on a probability space, with . Suppose and almost surely, for a common . With ,
At a zero denominator, the right-hand side is for and zero for . This is a one-sided result; a bound on the full spectral norm of each is not assumed.
import Definitions.Def_TroppMatrixConcentration_probability import Definitions.Def_TroppMatrixConcentration_dilation import Mathlib.Analysis.Convex.Function open MeasureTheory ProbabilityTheory open scoped Matrix.Norms.L2Operator
namespace TroppMatrixConcentration
theorem hermitian_bernstein {Ω : Type*} [MeasurableSpace Ω]
(μ : Measure Ω) [IsProbabilityMeasure μ] {d N : ℕ} [NeZero d]
(X : Fin N → Ω → Matrix (Fin d) (Fin d) ℂ) (L : ℝ) (hL : 0 ≤ L)
(hMeas : ∀ k, Measurable (X k))
(hHerm : ∀ k, ∀ᵐ ω ∂μ, (X k ω).IsHermitian)
(hL2 : ∀ k, MemLp (X k) 2 μ) (hIndep : iIndepFun X μ)
(hMean : ∀ k, (∫ ω, X k ω ∂μ) = 0)
(hBound : ∀ k, ∀ᵐ ω ∂μ, lambdaMax (X k ω) ≤ L) :
let Y := fun ω => ∑ k, X k ω
let v := hermitianSecondMoment μ Y
v = spectralNorm (∑ k, ∫ ω, X k ω ^ 2 ∂μ) ∧
(∫ ω, lambdaMax (Y ω) ∂μ) ≤
Real.sqrt (2 * v * Real.log d) + L * Real.log d / 3 ∧
∀ t : ℝ, 0 ≤ t → (μ {ω | t ≤ lambdaMax (Y ω)}).toReal ≤
bernsteinTail d v L t := by sorry
end TroppMatrixConcentration
Read-back
What the Lean code literally says, in plain math · gpt-6
Let be any probability space, let and be integers, and let , for , be measurable, mutually independent matrix-valued functions such that each is Hermitian almost surely, belongs to for the Euclidean operator norm, and has zero matrix-valued Bochner integral. For any square complex matrix , define , with the supremum of the empty set defined to be ; for Hermitian matrices this is the largest eigenvalue. Suppose satisfies and, for each , almost surely. Set and , where squares are matrix products and is the operator norm induced by the standard Euclidean norms. Then all of the following hold: ; , where is the natural logarithm; and, for every real , , with the probability viewed as a real number and with if and , if and , and otherwise. The quantifiers allow , in which case the assumptions on the family are vacuous and and , and allow , , , and ; for the bound is always , and for the upper bound on the expected largest eigenvalue is .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.