bernoulli_sampled_row_column_energy_controlled_log_moment_bound_2pN
Provedconcentrationkhintchinematrix-completion
Strengthened one-sample controlled row/column sampled-energy log-moment estimate (_2pN window). Under the Candes-Recht one-sample lower bound (, , ), there is a universal constant and an exponent with , , AND the two re-exported window bounds and (where ), such that the Bernoulli expectation of the -th power of the maximum sampled row/column energy is at most . This is the strengthening of node 4337294c that re-exports the exponent window the noncommutative-Khintchine conversion consumes.
Preamble
import Definitions.Def_matrix_completion_neumann import Definitions.Def_matrix_completion_tangent import Definitions.Def_matrix_completion_sampled_counts open MatrixCompletion
Formal statement
theorem bernoulli_sampled_row_column_energy_controlled_log_moment_bound_2pN :
∃ C : ℝ, 0 < C ∧
∀ (β : ℝ), 2 < β →
∀ (n₁ n₂ m : ℕ) (X : Matrix (Fin n₁) (Fin n₂) ℝ),
0 < n₁ → 0 < n₂ → m ≤ n₁ * n₂ →
(m : ℝ) ≥ β * (↑(max n₁ n₂)) *
Real.log (↑(max n₁ n₂)) →
2 ≤ max n₁ n₂ →
∃ q : ℕ, 1 ≤ q ∧
(q : ℝ) ≥ β * Real.log (↑(max n₁ n₂)) ∧
(q : ℝ) ≤ 2 * (β * Real.log (↑(max n₁ n₂))) ∧
(q : ℝ) ≤ 2 * (((m : ℝ) / ((n₁ : ℝ) * (n₂ : ℝ))) *
(↑(max n₁ n₂))) ∧
bernoulliExpectation ((m : ℝ) / ((n₁ : ℝ) * (n₂ : ℝ)))
(fun Omega =>
(max (sampledRowEnergyMax Omega X)
(sampledColumnEnergyMax Omega X)) ^ q) ≤
(C * ((m : ℝ) / ((n₁ : ℝ) * (n₂ : ℝ))) *
(↑(max n₁ n₂)) * entrySupNorm X ^ 2) ^ q := by
sorrySource
Candes-Recht 2008 (arXiv:0805.4471) Section 6.1 Lemma 6.2