Theorem 7.3.1 — Intrinsic matrix Bernstein
ProvedTroppMatrixConcentration.ch7_intrinsic_bernsteinLet be a finite independent family of measurable complex random matrices on a probability space, , with and almost surely, . Set . Let and , with at least one nonzero, and put , . Then
The variance bounds are matrices, not merely scalar norm bounds. The nonzero block proxy keeps intrinsic dimension in the source's meaningful domain. The finite family may be empty, and deterministic zero summands remain permitted. All required moments follow from boundedness. This is the designated goal; Theorem 7.2.1 is the other main theorem in this chapter proposal.
import Definitions.Def_TroppMatrixConcentration_ch7_intrinsic open MeasureTheory ProbabilityTheory open scoped Matrix.Norms.L2Operator ComplexOrder
namespace TroppMatrixConcentration
theorem ch7_intrinsic_bernstein {Ω : Type*} [MeasurableSpace Ω]
(μ : Measure Ω) [IsProbabilityMeasure μ] {m n N : ℕ} [NeZero m] [NeZero n]
(S : Fin N → Ω → Matrix (Fin m) (Fin n) ℂ) (L : ℝ) (hL : 0 ≤ L)
(V₁ : Matrix (Fin m) (Fin m) ℂ) (V₂ : Matrix (Fin n) (Fin n) ℂ)
(hV : V₁ ≠ 0 ∨ V₂ ≠ 0)
(hMeas : ∀ k, Measurable (S k)) (hIndep : iIndepFun S μ)
(hMean : ∀ k, (∫ ω, S k ω ∂μ) = 0)
(hBound : ∀ k, ∀ᵐ ω ∂μ, spectralNorm (S k ω) ≤ L)
(hVariance₁ : loewnerLE
(∫ ω, (∑ k, S k ω) * (∑ k, S k ω).conjTranspose ∂μ) V₁)
(hVariance₂ : loewnerLE
(∫ ω, (∑ k, S k ω).conjTranspose * (∑ k, S k ω) ∂μ) V₂) :
let Z := fun ω => ∑ k, S k ω
let r := intrinsicDimension (Matrix.fromBlocks V₁ 0 0 V₂)
let v := max (spectralNorm V₁) (spectralNorm V₂)
(∫ ω, Z ω * (Z ω).conjTranspose ∂μ) =
∑ k, ∫ ω, S k ω * (S k ω).conjTranspose ∂μ ∧
(∫ ω, (Z ω).conjTranspose * Z ω ∂μ) =
∑ k, ∫ ω, (S k ω).conjTranspose * S k ω ∂μ ∧
∀ t : ℝ, Real.sqrt v + L / 3 ≤ t →
(μ {ω | t ≤ spectralNorm (Z ω)}).toReal ≤
4 * r * Real.exp (-(t ^ 2 / 2) / (v + L * t / 3)) := by sorry
end TroppMatrixConcentration
Read-back
What the Lean code literally says, in plain math · gpt-6
For every measurable space , every probability measure on , every positive natural numbers , every natural number (including ), every family of measurable, mutually independent random matrices indexed by (with the Borel measurable structure on matrices and independence taken with respect to ), every real number , and every pair of matrices and with or , assume that each matrix-valued Bochner integral equals zero and that for -almost every , separately for every . Write , and further assume that both and are positive semidefinite, meaning Hermitian with nonnegative quadratic forms; here is the conjugate transpose and is the operator norm for the Euclidean norms. Set , , and . Then all three of the following hold: ; ; and, for every real satisfying , the probability, converted from an extended nonnegative real to a real number, satisfies . When , all indexed hypotheses are vacuous and all sums are zero; the same nonzero requirement on the pair and the two positive semidefinite difference hypotheses still apply. The case is included. The nonzero requirement ensures and , so the displayed divisions have nonzero denominators for every permitted .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.