Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 8 — correspondence preserves average optimality

Proved
HordijkKallenbergLP.Correspondence.theorem_8

by mikedeng1 · Oct 6, 2026 · Mathlib 0df444a (Lean v4.33.1)

average-rewardlinear-programmingmarkov-decision-processmultichainp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1

Let βi>0\beta_i>0βi​>0 with ∑iβi=1\sum_i\beta_i=1∑i​βi​=1 in a finite Markov decision process. There are two directions.

  1. If a stationary randomized policy π∞\pi^\inftyπ∞ is average optimal among all history-dependent randomized policies, then its representative (x(π),y(π))(x(\pi),y(\pi))(x(π),y(π)) from equation (6) is an optimal solution of the dual linear program.
  2. If (x,y)(x,y)(x,y) is any optimal dual solution, then its normalized stationary randomized policy π∞(x,y)\pi^\infty(x,y)π∞(x,y) exists and is average optimal.
π∞ average optimal⟹(x(π),y(π)) dual optimal,(x,y) dual optimal⟹π∞(x,y) average optimal.\pi^\infty\text{ average optimal}\Longrightarrow (x(\pi),y(\pi))\text{ dual optimal},\qquad (x,y)\text{ dual optimal}\Longrightarrow\pi^\infty(x,y)\text{ average optimal}.π∞ average optimal⟹(x(π),y(π)) dual optimal,(x,y) dual optimal⟹π∞(x,y) average optimal.

The theorem relates policy performance to every dual optimum in a multichain model. Formalization Note The second direction explicitly constructs a randomized policy with the printed weights; no extreme-point or pure-action hypothesis is imposed.

Preamble
import Mathlib
import Definitions.Def_HordijkKallenbergLP_Correspondence_Representative
Formal statement
namespace HordijkKallenbergLP.Correspondence

open MarkovDecisionProcesses

/-- Hordijk and Kallenberg (1979), Theorem 8(a) and (b), p. 360.
Formalization Note: part (b) asserts existence of the normalized randomized
policy with the exact printed weights and its average optimality. It ranges over
all dual optima, not only extreme points. -/
theorem theorem_8 {S A : Type*} [Fintype S] [Fintype A]
    [DecidableEq S] [DecidableEq A] [Nonempty S]
    (M : StationaryMDP S A) (β : S → ℝ)
    (hβpos : ∀ i, 0 < β i) (hβsum : ∑ i, β i = 1) :
    (∀ π : RandomizedPolicy M, HordijkKallenbergLP.SingleLP.IsAvgOptimal M π.toHR →
      DualOptimal M β (representativeX M β π) (representativeY M β π)) ∧
    (∀ x y : HordijkKallenbergLP.SingleLP.Pair M → ℝ, DualOptimal M β x y →
      ∃ π : RandomizedPolicy M,
        (∀ i a, π.weight i a = dualPolicyWeight M x y i a) ∧
        HordijkKallenbergLP.SingleLP.IsAvgOptimal M π.toHR) := by sorry

end HordijkKallenbergLP.Correspondence
Source
Hordijk and Kallenberg, Linear Programming and Markov Decision Chains, Management Sci. 25(4):352–362 (1979), DOI 10.1287/mnsc.25.4.352, p. 360, Theorem 8
Human review
  • Endorsed by Shuze Chen · Oct 6, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 6, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me