InformationTheory

InformationTheory.Shannon.EPI.G2.KLVariationalLower

source

Donsker–Varadhan variational lower bound (easy direction) #

The variational lower bound on KL divergence: KL(μ ‖ ν) ≥ ∫ g dμ − log (∫ exp(g) dν) for all bounded measurable g.

Main statements #

Implementation notes #

The proof applies Jensen's inequality (ConvexOn.map_integral_le) to h := g − llr μ ν, then uses the change-of-measure identity (integral_toReal_rnDeriv_mul) to push to ν. The hard direction of Donsker–Varadhan (sup attainment) is not in scope here.

theorem

InformationTheory.Shannon.integral_exp_sub_llr_le

source
{α : Type u_1} { : MeasurableSpace α} {μ ν : MeasureTheory.Measure α} [MeasureTheory.IsProbabilityMeasure μ] [MeasureTheory.IsProbabilityMeasure ν] (hμν : μ.AbsolutelyContinuous ν) {g : α} (hg_meas : Measurable g) {C : } (hg_bdd : ∀ (x : α), |g x| C) :
(x : α), Real.exp (g x - MeasureTheory.llr μ ν x) μ (x : α), Real.exp (g x) ν

Change-of-measure inequality: when μ ≪ ν, ∫ exp (g x − llr μ ν x) ∂μ ≤ ∫ exp (g x) ∂ν for all bounded measurable g. @audit:ok

Used by
    theorem

    InformationTheory.Shannon.klDiv_variational_lower_bound

    source
    {α : Type u_1} { : MeasurableSpace α} {μ ν : MeasureTheory.Measure α} [MeasureTheory.IsProbabilityMeasure μ] [MeasureTheory.IsProbabilityMeasure ν] (hμν : μ.AbsolutelyContinuous ν) (h_int : MeasureTheory.Integrable (MeasureTheory.llr μ ν) μ) {g : α} (hg_meas : Measurable g) {C : } (hg_bdd : ∀ (x : α), |g x| C) :
    (x : α), g x μ - Real.log ( (x : α), Real.exp (g x) ν) (klDiv μ ν).toReal

    Donsker–Varadhan variational lower bound (easy direction): ∫ g ∂μ − log (∫ exp g ∂ν) ≤ (klDiv μ ν).toReal for all bounded measurable g. @audit:ok

    Used by