Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 3.2, p. 7 — the mapping of an averager is a max-norm nonexpansion, hence compatible with discounted value backups

Proved
StableFA.Discounted.theorem_3_2

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

averagerfunction-approximationmax-normnonexpansionp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1value-iteration

Let AAA be an averager on nnn states, with constants kik_iki​ and nonnegative weights βi,βij\beta_i,\beta_{ij}βi​,βij​ satisfying βi+∑jβij=1\beta_i+\sum_j\beta_{ij}=1βi​+∑j​βij​=1, and let MA(Y)i=βiki+∑jβijYjM_A(Y)_i=\beta_ik_i+\sum_j\beta_{ij}Y_jMA​(Y)i​=βi​ki​+∑j​βij​Yj​ be its mapping. Then:

  1. MAM_AMA​ is a nonexpansion in the max norm: for all Y,Z∈RnY,Z\in\mathbb R^nY,Z∈Rn,
∥MA(Y)−MA(Z)∥≤∥Y−Z∥;\|M_A(Y)-M_A(Z)\|\le\|Y-Z\| ;∥MA​(Y)−MA​(Z)∥≤∥Y−Z∥;
  1. MAM_AMA​ is compatible with the parallel value backup operator TMT_MTM​ of every discounted finite MDP MMM on the same nnn states (discount 0≤γ<10\le\gamma<10≤γ<1, stochastic transition rows): for every initial guess V0V_0V0​ the iterates (MA∘TM)k(V0)(M_A\circ T_M)^k(V_0)(MA​∘TM​)k(V0​) converge.

The theorem singles out the averagers as a class of function approximators that cannot exaggerate differences between target value functions, and therefore can be combined with value iteration without risking divergence.

Formalization Note ∥⋅∥\|\cdot\|∥⋅∥ on Fin n → ℝ is the max norm. The MDP is BertsekasSSPModel with the stochastic-row hypothesis, any finite action type and any nonempty admissible action sets. The second part is stated for every such MDP, as on the page ("for any discounted MDP").

Preamble
import Mathlib
import Definitions.Def_BertsekasSSPModel
import Definitions.Def_StableFA_Discounted_Setting

open Filter Topology
Formal statement
namespace StableFA.Discounted

theorem theorem_3_2 {n : ℕ} (A : Averager n) :
    (∀ Y Z : Fin n → ℝ, ‖A.apply Y - A.apply Z‖ ≤ ‖Y - Z‖) ∧
    ∀ (C : Type) [Fintype C] (M : BertsekasSSPModel n C) (γ : ℝ), 0 ≤ γ → γ < 1 →
      (∀ i, ∀ u ∈ M.U i, ∑ j, M.p i u j = 1) →
      Compatible A.apply (BertsekasDiscountedBellmanOp M γ) := by sorry

end StableFA.Discounted
Source
Gordon, Stable Function Approximation in Dynamic Programming, Tech. Rep. CMU-CS-95-103, Carnegie Mellon University (1995), p. 7, Theorem 3.2
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