Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Lemma 7, p. 13 (FPL* with perturbation −R/ε; sign corrected) — E[Switches_τ(FPL*_ε)] ≤ (ετ + 2 log n)/(1 − ε)

Proved
LimitedSwitching.HighProb.lemma_7

by mikedeng1 · Oct 19, 2026 · Mathlib 0df444a (Lean v4.33.1)

follow-the-perturbed-leaderonline-learningp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1switching

Let n≥1n\ge1n≥1 and 0<ε<10<\varepsilon<10<ε<1, and let (ℓt)(\ell_t)(ℓt​) be any loss sequence with values in [0,1][0,1][0,1], fixed in advance. Run FPLε∗^*_\varepsilonε∗​: draw R(1),…,R(n)R(1),\dots,R(n)R(1),…,R(n) i.i.d. standard exponential, and at each round ttt play the action minimising −R(i)/ε+∑s=1t−1ℓs(i)-R(i)/\varepsilon+\sum_{s=1}^{t-1}\ell_s(i)−R(i)/ε+∑s=1t−1​ℓs​(i). Then for every horizon τ∈N\tau\in\mathbb Nτ∈N the expected number of switches satisfies

E[Switchesτ(FPLε∗)]≤11−ε (ετ+2log⁡n).\mathbb E\big[\mathrm{Switches}_\tau(\mathrm{FPL}^*_\varepsilon)\big]\le\frac{1}{1-\varepsilon}\,(\varepsilon\tau+2\log n).E[Switchesτ​(FPLε∗​)]≤1−ε1​(ετ+2logn).

This is property (i) of Framework 1 for FPL∗^*∗. Through Lemma 3 it gives the high-probability switching bound of BFPLδ∗^*_\deltaδ∗​.

Formalization Note The paper's FPLε∗^*_\varepsilonε∗​ has P1(i)=+R(i)/εP_1(i)=+R(i)/\varepsilonP1​(i)=+R(i)/ε. For that variant the bound fails numerically (with n=1000n=1000n=1000, ε=0.01\varepsilon=0.01ε=0.01, τ=200\tau=200τ=200 it averages about 25.8 switches against a bound of 16.0). The statement is made for the sign-corrected perturbation −R(i)/ε-R(i)/\varepsilon−R(i)/ε, the exponential analogue of Kalai–Vempala's algorithm. The hypothesis ε<1\varepsilon<1ε<1 makes 1/(1−ε)1/(1-\varepsilon)1/(1−ε) meaningful. The expectation is a lower Lebesgue integral of the natural-valued switch count.

Preamble
import Mathlib
import Definitions.Def_LimitedSwitching_HighProb_Setting

open MeasureTheory ProbabilityTheory
open scoped ENNReal
Formal statement
namespace LimitedSwitching.HighProb

theorem lemma_7 {n : ℕ} [NeZero n] (ε : ℝ) (hε : 0 < ε) (hε1 : ε < 1) (τ : ℕ)
    (ℓ : ℕ → Fin n → ℝ) (hℓ : IsLossSeq ℓ) :
    ∫⁻ R, (switches τ (play (fplStar ε) R ℓ) : ℝ≥0∞) ∂(expLaw n)
      ≤ ENNReal.ofReal ((ε * τ + 2 * Real.log n) / (1 - ε)) := by sorry

end LimitedSwitching.HighProb
Source
Altschuler & Talwar, Online learning over a finite action set with limited switching, arXiv:1803.01548v2, p. 13, Lemma 7
Human review
  • Endorsed by Shuze Chen · Oct 19, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 19, 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