Derivative of a Tilted Covariance
ProvedFunctionalInequalities.Probability.Posterior.covariance_hasDerivAt_of_compact_modularfunctional-inequalitieskls-conjecturekls-modularlog-concave
This is the compact-support derivative formula for a tilted covariance.
Let be a compactly supported probability measure on a topological, measurable, second-countable Borel space. Let be continuous and let be measurable and bounded on by some . Then
where . This provides the covariance derivative needed for second-order posterior calculations.
Superseded upload. This snapshot was replaced by FunctionalInequalities.Probability.Posterior.covariance_hasDerivAt_of_compact_readable: open the readable replacement. The replacement preserves the frozen source and mathematical statement, with a compact declaration context. This historical node remains available for existing imports.
Preamble
import Definitions.Def_KLSModular_FunctionalInequalities_Probability_Posterior_CompactPrior
noncomputable section
open MeasureTheory Filter Topology ProbabilityTheory
open scoped RealInnerProductSpace
section
namespace FunctionalInequalities
end FunctionalInequalities
open FunctionalInequalities
section
namespace FunctionalInequalities.Probability
end FunctionalInequalities.Probability
open FunctionalInequalities FunctionalInequalities.Probability
section
namespace FunctionalInequalities.Probability.Posterior
end FunctionalInequalities.Probability.Posterior
open FunctionalInequalities FunctionalInequalities.Probability FunctionalInequalities.Probability.Posterior
variable {Ω : Type*} [TopologicalSpace Ω] [MeasurableSpace Ω] [BorelSpace Ω]
[SecondCountableTopology Ω]
Formal statement
theorem FunctionalInequalities.Probability.Posterior.covariance_hasDerivAt_of_compact_modular (μ : ProbabilityMeasure Ω)
(hc : IsCompact (μ : Measure Ω).support) {φ g k f : Ω → ℝ}
(hφ : Continuous φ) (hg : Continuous g) (hk : Continuous k) (hf : Measurable f)
{C : ℝ} (hC : 0 ≤ C) (hfb : ∀ x ∈ (μ : Measure Ω).support, ‖f x‖ ≤ C) :
HasDerivAt (fun t : ℝ => covariance f k
((μ : Measure Ω).tilted (fun x => φ x + t * g x)))
(covariance f (fun x => g x * k x) ((μ : Measure Ω).tilted φ) -
(∫ x, g x ∂((μ : Measure Ω).tilted φ)) * covariance f k ((μ : Measure Ω).tilted φ) -
(∫ x, k x ∂((μ : Measure Ω).tilted φ)) * covariance f g ((μ : Measure Ω).tilted φ)) 0 := by sorry
Source
Lean source: https://github.com/human-ai-mathematics/kls-lean/blob/408915f/formalization/FunctionalInequalities/Probability/Posterior/CompactPrior.lean#L202-L260 (repository human-ai-mathematics/kls-lean, commit 408915f, module `FunctionalInequalities.Probability.Posterior.CompactPrior`, declaration `FunctionalInequalities.Probability.Posterior.covariance_hasDerivAt_of_compact`; the repository is private at upload time).