Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 5.4 — unprojected OSD with t^(−α) steps can incur order T^(2−α) regret

Proved
ModernOnlineLearning.LowerBounds.theorem_5_4

by mikedeng1 · Oct 9, 2026 · Mathlib 0df444a (Lean v4.33.1)

lower-boundsonline-learningp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-bookp2o-v1subgradient-descent

Fix a Euclidean dimension d≥1d\ge1d≥1, 0<α<10<\alpha<10<α<1, and define ϕ(α)\phi(\alpha)ϕ(α) as above. If the horizon satisfies T≥2/((1−α)ϕ(α))T\ge2/((1-\alpha)\phi(\alpha))T≥2/((1−α)ϕ(α)), then there is a sequence of convex, 111-Lipschitz losses for which unprojected online subgradient descent, initialized at x1=0x_1=0x1​=0 and using step size t−αt^{-\alpha}t−α, has

Regret⁡T(0)≥12ϕ(α)T2−α.\operatorname{Regret}_T(0)\ge\tfrac12\phi(\alpha)T^{2-\alpha}.RegretT​(0)≥21​ϕ(α)T2−α.

This shows that the same decaying steps can perform much worse when there is no bounded feasible domain.

Formalization Note The losses are represented by linear functions z↦⟨gt,z⟩z\mapsto\langle g_t,z\ranglez↦⟨gt​,z⟩ with ∥gt∥2≤1\|g_t\|_2\le1∥gt​∥2​≤1, as in the book's construction. Their convexity and Lipschitz property follow from this form. The displayed limit lim⁡α→1ϕ(α)=1−ln⁡2\lim_{\alpha\to1}\phi(\alpha)=1-\ln2limα→1​ϕ(α)=1−ln2 is a separate claim and is outside this item.

Preamble
import Mathlib
import Definitions.Def_ModernOnlineLearning_LowerBounds_phi
Formal statement
namespace ModernOnlineLearning.LowerBounds

/-- The existence and regret-bound part of Theorem 5.4, printed p. 52.
The adversarial losses are linear along one coordinate, as in the source proof;
the limit statement about `φ` is not part of this declaration. -/
theorem theorem_5_4 {d : ℕ} (hd : 0 < d)
    (α : ℝ) (hαpos : 0 < α) (hαlt : α < 1)
    (T : ℕ) (hT : 2 / ((1 - α) * phi α) ≤ (T : ℝ)) :
    ∃ (g x : ℕ → EuclideanSpace ℝ (Fin d)),
      x 1 = 0 ∧
      (∀ t ∈ Finset.Icc 1 T, ‖g t‖ ≤ 1) ∧
      (∀ t ∈ Finset.Icc 1 T,
        x (t + 1) = x t - ((t : ℝ) ^ (-α)) • g t) ∧
      (1 / 2 : ℝ) * phi α * (T : ℝ) ^ (2 - α) ≤
        ∑ t ∈ Finset.Icc 1 T, inner ℝ (g t) (x t) := by sorry

end ModernOnlineLearning.LowerBounds
Source
Orabona, arXiv:1912.13213v10, Theorem 5.4 and its construction, p. 52
Human review
  • Endorsed by Shuze Chen · Oct 9, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 9, 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