2025/04/15 by Weber, Michel · 1 citation
#42A38 #60G50 #FOS: Mathematics #Probability (math.PR)
paper · doi:10.48550/arxiv.2504.11198
We use decoupling inequalities for general Gaussian vectors and classical tools from Gaussian processes theory to study moderate deviations of suprema of trigonometric and almost periodic Gaussian polynomials. Let Xx(t) amp;=amp; ∑k≤ x ak ( gk cos 2πk t+ g'ksin 2πk t ), 0≤ t≤ 1, xgt;0, where (gk)k≥ 1, (g'k)k≥ 1 are two independent sequences of i.i.d. \mathcal N(0,1) distributed random variables, and (ak)k≥ 1 are real numbers such that Ax = ∑k≤ x ak2\uparrow ∞ with x. Assume for instance that A(x)∼ loglog x, x→ ∞ and B=∑k≥ 1 ak4<∞. Let 0<η< 1. We prove that there exists an absolute constant C, such that for all x large enough, ¶\ sup0≤ t≤ 1 Xx(t) ≤√(2η(loglog x)(log loglog x))\ ≤ e^- \fracC (loglog x)1-η √(8η(B+1)(log loglog x)). We also establish an approximation theorem of moderate deviations of almost periodic Gaussian polynomials by Gaussian polynomials with \Q-frequencies, and show that the error term has an exponential decay. We finally study for general non-vanishing coefficient sequences, the behavior along lattices of almost periodic Gaussian polynomials with linearly independent frequencies.