Alphabeta Math
TheoremStatement: Literature-sourcedProof: AI-adaptedPipeline-generated
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

Finite-alphabet conditional entropy is version-invariant and obeys its chain rule

Statement

Assume AC. Let X,U take values in finite nonempty sets and Z in a standard-Borel space. Conditional entropy as defined in Finite-alphabet conditional entropy given an observation is unchanged when any supplied regular conditional law is replaced by another version. Moreover, H(X,U∣Z)=H(X∣Z)+H(U∣X,Z). All three quantities are finite; the identity is an additive finite-sum identity.

Facts & Assumptions

Given: The random elements and hypotheses in the Statement.

[A1]

AC is required for the stated regular conditional law existence and simultaneous uniqueness interfaces. The Axiom of Choice

[F1]

H(X∣Z) is the integral of the finite Shannon entropy of a conditional probability vector; the value of −tln⁡t at zero is zero. Finite-alphabet conditional entropy given an observation Finite distributions, entropy, divergence, and stochastic matrices

[F2]

Two regular conditional distributions of a standard-Borel-valued variable agree as probability measures outside one measurable null set. Simultaneous ae uniqueness of regular conditional distributions

[F3]

A joint law on standard-Borel factors disintegrates through a supplied regular conditional law, and the integral identity extends from rectangles to nonnegative product-measurable functions. Disintegration of a joint law on standard borel spaces

[F4]

A finite product of standard-Borel spaces is standard Borel. Finite products of standard Borel spaces are standard Borel

Proof

Given: AC and the stated finite-valued random elements and standard-Borel conditioning variable.

1.1A1F1F2given

If two versions of the conditional law of X given Z are supplied, [F2] gives one PZ-null set outside which their probability measures agree. Every atom probability, and hence the finite entropy sum in [F1], then agrees pointwise outside that set. Integrating gives the same H(X∣Z). The same argument applies to (X,U)∣Z and to U∣(X,Z).

1.2A1F2F3F4givenconstruct

Choose regular conditional laws of (X,U) given Z and of U given (X,Z); they exist by [A1, F3], since (X,U) and (X,Z) are standard Borel by [F4]. Write pij(z)=P(X=i,U=j∣Z=z), pi(z)=∑jpij(z), and mij(z)=P(U=j∣X=i,Z=z) for the chosen versions. The function pi is itself a conditional law of X given Z, so [F2] makes it agree almost everywhere with the version used in H(X∣Z). For each i,j and measurable C in the value space of Z, the two disintegrations [F3] give ∫Cpi(z)mij(z) PZ(dz)=P(X=i,U=j,Z∈C)=∫Cpij(z) PZ(dz). Both integrands are bounded and measurable. Testing the sets where one exceeds the other by at least 1/n, for each n≥1, shows pi(z)mij(z)=pij(z) almost everywhere; because the alphabets are finite, one null set works for every i,j.

2.1F1step 1.2algebra

Outside that null set, if pi(z)=0 then every pij(z)=0, so the row contributes zero to the conditional entropy. If pi(z)>0, then mij(z)=pij(z)/pi(z) for every j. On positive entries, ln⁡pij=ln⁡pi+ln⁡mij; multiplying by −pij and summing over the finite alphabets gives H(X,U∣Z=z)=H(X∣Z=z)+∑ipi(z)H(U∣X=i,Z=z).

3.1F1F3step 1.1step 2.1algebra∎

Integrate the finite conditional identity from step 2.1 with respect to PZ. The disintegration identity [F3] identifies the last integral with H(U∣X,Z) and the other terms with their definitions. By the bound −tln⁡t≤1/e in [F1], each conditional entropy is bounded by its finite alphabet size divided by e, so all integrals are finite and no subtraction of infinite quantities occurs. Step 1.1 proves the accompanying version-independence assertion.

Depends on

Used by

Nothing in the library uses this result yet.

Dependency tree · two levels

47 results within two dependency steps of this one, each drawn at its shortest distance from it. An arrow runs from a result to what uses it, so the chart reads left to right and ends at this result, which carries a heavier outline. Every node is a link to that result. Click elsewhere on the chart to enlarge it.

Sources