Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proof of Theorem 10, p. 362 — by definition of γ, ȳ(f) vanishes at some state of every ergodic set of P(f)

Proved
HordijkKallenbergLP.ExtremePoint.repY_zero_on_ergodic

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

linear-programmingmarkov-chainmarkov-decision-processp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1

Let βj>0\beta_j>0βj​>0 with ∑jβj=1\sum_j\beta_j=1∑j​βj​=1, let f∞f^\inftyf∞ be a pure stationary policy (f(i)∈A(i)f(i)\in A(i)f(i)∈A(i)), and let yˉi(f)=yif(i)(f)\bar y_i(f)=y_{if(i)}(f)yˉ​i​(f)=yif(i)​(f) be the yyy-part of its representative (6) read along fff. For every ergodic set EkE_kEk​ of P(f)P(f)P(f) there is a state i(k)∈Eki(k)\in E_ki(k)∈Ek​ with

yˉi(k)(f)=0.\bar y_{i(k)}(f)=0.yˉ​i(k)​(f)=0.

This zero is what the choice of γ\gammaγ in (6) is for: it anchors the uniqueness argument on each ergodic set in the proof of Theorem 10.

Formalization Note. Ergodic sets are the sets of states accessible from a recurrent state of P(f)P(f)P(f). The vector γ\gammaγ uses the denominator ∑k∈Ejpki∗\sum_{k\in E_j}p^*_{ki}∑k∈Ej​​pki∗​; with the printed ∑kpki∗\sum_k p^*_{ki}∑k​pki∗​ the zero need not exist.

Preamble
import Mathlib
import Definitions.Def_MarkovDecisionProcesses_AverageReward
import Definitions.Def_BlackwellDiscreteDP_NearOne_LimitMatrix
import Definitions.Def_HordijkKallenbergLP_ExtremePoint_Model
open Matrix MarkovDecisionProcesses BlackwellDiscreteDP.NearOne
Formal statement
namespace HordijkKallenbergLP.ExtremePoint

/-- **Proof of Theorem 10, p. 362 (unnumbered): a zero of `ȳ(f)` on every ergodic set.** Let
`f^∞` be a pure stationary policy (`f(i) ∈ A(i)`) and `P = P(f)`. On every ergodic set `E_k`
of `P` there is a state `i(k)` with `ȳ_{i(k)}(f) = y_{i(k) f(i(k))}(f) = 0`, where `y(f)` is the
`y`-part of the representative (6).

Hordijk and Kallenberg, Linear Programming and Markov Decision Chains, Management Sci.
25(4):352–362 (1979), DOI 10.1287/mnsc.25.4.352, p. 362, proof of Theorem 10
("By definition of γ, there is on each E_k a state i(k) such that ȳ_{i(k)}(f) = 0;").

**Formalization Note.** The ergodic sets are the sets `ergodicSet P l` of states accessible from a
recurrent `l`; `P(f)` is `policyMatrix M (pureRule f)`, which equals
`(p_{if(i)j})` = `transMatrix M f`. `γ` uses the corrected denominator `∑_{k ∈ E_j} p*_ki`
(see `gamma`); with the printed `∑_k p*_ki` this zero need not exist. -/
theorem repY_zero_on_ergodic {S A : Type*} [Fintype S] [Fintype A] [DecidableEq S]
    [DecidableEq A] (M : StationaryMDP S A) (β : S → ℝ) (hβpos : ∀ j, 0 < β j)
    (hβsum : ∑ j, β j = 1) (f : S → A) (hf : ∀ i, f i ∈ M.admissible i) (l : S)
    (hl : IsRecurrent (policyMatrix M (pureRule f)) l) :
    ∃ i ∈ ergodicSet (policyMatrix M (pureRule f)) l,
      repY M β (pureRule f) ⟨(i, f i), hf i⟩ = 0 := by sorry

end HordijkKallenbergLP.ExtremePoint
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. 362, proof of Theorem 10
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