Lemma 5 — matrix induced norm bounded by twice the trace
ProvedDiscreteChoiceProx.Simplex.lemma_5convex-conjugatediscrete-choicematrix-normp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1prox-functionrandom-utility
Let be a symmetric real matrix in the paper's class : its diagonal entries are nonnegative, off-diagonal entries are nonpositive, and each row sums to zero. Then
Here . The bound is the matrix estimate used in the proof of Theorem 2. It is stated for every such finite matrix, independently of any random utility law.
Preamble
import Mathlib import Definitions.Def_DiscreteChoiceProx_Simplex_Model
Formal statement
namespace DiscreteChoiceProx.Simplex
/-- Lemma 5, p. 11: the induced norm of a matrix in `𝒜`. -/
theorem lemma_5 {n : ℕ} [NeZero n] (A : Matrix (Fin n) (Fin n) ℝ) (hA : ClassA A) :
normInfOne A ≤ 2 * A.trace := by sorry
end DiscreteChoiceProx.Simplex
Source
Müller, Nesterov, Shikhman, Discrete choice prox-functions on the simplex, arXiv:1909.05591v1, p. 11, Lemma 5
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.