OnlineLearning.exists_forall_le_of_finite_sublevel_intersections
theoremverifiedimported Lean
From finite feasibility to a uniform row point. Let $X, Y \subseteq \bbr$ together with a payoff function $f : \bbr \to \bbr \to \bbr$ satisfy the convex-compact minimax hypotheses, and let $c \in \bbr$. Suppose that for every finite subset $u \subseteq Y$ the set $X \cap \bigcap_{y \in u} \mathrm{sublevel}(X,f,y,c)$ is nonempty, where $\mathrm{sublevel}(X,f,y,c) = X \cap \{x \in \bbr : f(x,y) \le c\}$; that is, there is a common row point $x \in X$ with $f(x,y) \le c$ for all $y$ in that finite set. Then there exists a single $x \in X$ with $f(x,y) \le c$ for every $y \in Y$.
Statement
set_option relaxedAutoImplicit false in set_option autoImplicit false in open Real Finset BigOperators in lemma OnlineLearning.exists_forall_le_of_finite_sublevel_intersections {X Y : Set ℝ} {f : ℝ → ℝ → ℝ} (h : ConvexCompactMinimaxHypotheses X Y f) (c : ℝ) (hfin : ∀ u : Finset Y, (X ∩ ⋂ y ∈ u, minimaxSublevel X f y c).Nonempty) : ∃ x ∈ X, ∀ y : Y, f x y ≤ c
- Source
- TCSlib, `OnlineLearning.exists_forall_le_of_finite_sublevel_intersections`
- Lean proof
- By Karim Abdel Sadek, Mark Bedaywi, from github.com/Shilun-Allan-Li/tcslib at
c8a02591b1e5(Copyright (c) 2026 Karim Abdel Sadek and Mark Bedaywi. All rights reserved, Apache-2.0). Compiled here against leanprover/lean4:v4.33.1, Mathlib v4.33.1; written for leanprover/lean4:v4.25.0.OnlineLearning.exists_forall_le_of_finite_sublevel_intersections← OnlineLearning.exists_forall_le_of_finite_sublevel_intersections (TCSlib/LearningTheory/Minimax/ConvexMinimaxCore.lean:412): verbatim
- Verified
- 07 Oct 2026
- Axioms
Classical.choiceQuot.soundpropext- Built on 3
- OnlineLearning.minimaxSublevel_isClosedClosedness of the minimax sublevel set. Let $X \subseteq \bbr$ be a closed set, let $f\colon \bbr \times \bbr \to \bbr$ be a payoff…
- OnlineLearning.minimaxSublevelFor a set $X \subseteq \bbr$, a payoff function $f$, a column point $y$, and a threshold $c \in \bbr$, the minimax sublevel set is…
- OnlineLearning.ConvexCompactMinimaxHypothesesA structure bundling the assumptions of Cesa-Bianchi--Lugosi Theorem~7.1 for $f : \bbr \to \bbr \to \bbr$ on sets $X, Y \subseteq \bbr$:…
- Used by 1
- OnlineLearning.convex_compact_minimax_of_finite_sublevel_intersectionsMinimax equality from finite sublevel intersections. Let $X, Y \subseteq \bbr$ with $X$ and $Y$ nonempty, $X$ compact and convex, and $Y$…
Lean source view module on GitHub
Choose a tactic — a marked line — to see the goals Lean had before and after it, as in Lean’s infoview. Step with ← → or j k.
set_option relaxedAutoImplicit false inset_option autoImplicit false inopen Real Finset BigOperators in/-- The finite-intersection compactness step: under `ConvexCompactMinimaxHypotheses`, if for every finite set `u` of column points there is a point of `X` with `f x y ≤ c` for all `y ∈ u`, then there is a single point `x ∈ X` with `f x y ≤ c` for every `y ∈ Y`. This is the finite intersection property of the compact set `X` applied to the closed sublevel sets `minimaxSublevel X f y c`. -/lemma OnlineLearning.exists_forall_le_of_finite_sublevel_intersections {X Y : Set ℝ} {f : ℝ → ℝ → ℝ} (h : ConvexCompactMinimaxHypotheses X Y f) (c : ℝ) (hfin : ∀ u : Finset Y, (X ∩ ⋂ y ∈ u, minimaxSublevel X f y c).Nonempty) : ∃ x ∈ X, ∀ y : Y, f x y ≤ c := by have hclosed : ∀ y : Y, IsClosed (minimaxSublevel X f y c) := by intro y exact minimaxSublevel_isClosed h.X_compact.isClosed (h.continuous_left y y.2) rcases h.X_compact.inter_iInter_nonempty (fun y : Y => minimaxSublevel X f y c) hclosed hfin with ⟨x, hx⟩ refine ⟨x, hx.1, ?_⟩ intro y -- Membership in every sublevel set is exactly the pointwise bound `f x y ≤ c`. exact (Set.mem_iInter.mp hx.2 y).2