Theorem 3.6.1 — Master expectation and tail bounds
ProvedTroppMatrixConcentration.master_boundsLet be a finite independent family of measurable integrable Hermitian random matrices on a probability space, with . Fix a real parameter such that every is integrable. Set and . If , then
If , then
Both tail bounds hold for every real . These are the pointwise-parameter forms of all four master bounds; optimization over admissible parameters gives the source's infima and supremum. The case imposes no conclusion.
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 master_bounds {Ω : Type*} [MeasurableSpace Ω]
(μ : Measure Ω) [IsProbabilityMeasure μ] {d N : ℕ} [NeZero d]
(X : Fin N → Ω → Matrix (Fin d) (Fin d) ℂ) (θ : ℝ)
(hMeas : ∀ k, Measurable (X k))
(hHerm : ∀ k, ∀ᵐ ω ∂μ, (X k ω).IsHermitian)
(hInt : ∀ k, Integrable (X k) μ) (hIndep : iIndepFun X μ)
(hExp : ∀ k, Integrable (fun ω => matrixExp (θ • X k ω)) μ) :
(0 < θ →
(∫ ω, lambdaMax (∑ k, X k ω) ∂μ) ≤
Real.log (traceExp (cumulantSum μ X θ)) / θ ∧
∀ t : ℝ, (μ {ω | t ≤ lambdaMax (∑ k, X k ω)}).toReal ≤
Real.exp (-θ * t) * traceExp (cumulantSum μ X θ)) ∧
(θ < 0 →
Real.log (traceExp (cumulantSum μ X θ)) / θ ≤
(∫ ω, lambdaMin (∑ k, X k ω) ∂μ) ∧
∀ t : ℝ, (μ {ω | lambdaMin (∑ k, X k ω) ≤ t}).toReal ≤
Real.exp (-θ * t) * traceExp (cumulantSum μ X θ)) := by sorry
end TroppMatrixConcentration
Read-back
What the Lean code literally says, in plain math · gpt-6
For every measurable space with a probability measure , every pair of natural numbers with , every family indexed by , and every real , assume that each is measurable for the usual Borel structure on matrices, is Hermitian -almost everywhere, and is Bochner integrable, that the family is jointly independent under (its generated sigma-algebras are independent), and that each function is Bochner integrable. Here matrix exponentiation is the norm-convergent series . Define , , and , where is the real continuous functional calculus applied to the real logarithm: for Hermitian it is , with and for , and it is the zero matrix for non-Hermitian . For an arbitrary square matrix , set , , and , with both values defined as when this real spectrum is empty; for Hermitian matrices these are their largest and smallest eigenvalues. Then both implications hold: if , and, for every , ; if , and, for every , . The measures in these inequalities are regarded as real numbers, which is possible because is a probability measure; the scalar integrals use the Bochner convention of value for a nonintegrable integrand. Hermitian values are required only almost everywhere, and the real-spectrum definitions also specify the expressions at exceptional points. There is no sign restriction on ; when , both antecedents are false and the conclusion imposes no inequalities (real division by zero is defined as ). The case is included, with the indexed assumptions vacuous, , , , and both extreme spectral values equal to ; is excluded.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.