betaFeedback

Reading v1View latest version

WorksMathematics

Gaussian Products: Deterministic Principal Ideals and Wick Order Recovery

Abstract

For fixed Hilbert–Schmidt operators A,BA,B and a standard Gaussian array GG, the algebraic principal ideal generated by AGBAGB almost surely equals a deterministic ideal obtained by multiplying the square tails of s(A)s(A) and s(B)s(B). In fixed coordinates, the corresponding singular-value comparisons hold simultaneously for all nonincreasing square-summable diagonal weights. We characterize the numerical arrays with this uniform realization property on any two prescribed rank flags by rectangular operator-norm upper bounds and Hilbert–Schmidt lower bounds; the optimal universal upper dilation is two. A family of strictly positive factors reflects every inclusion of principal ideals if and only if its tail-rank profiles have no common eventual upper bound. Otherwise, every local flag region contains a common fibre whose input ideals form a copy of P(N)/Fin\mathcal P(\mathbb N)/\mathrm{Fin} as a sublattice. For a specified family of operator-valued Gaussian chaoses, Wick composition with one fixed random operator yields principal ideals that recover these inclusion relations. The random ideal identities hold almost surely for each fixed parameter choice.

Introduction

Let A,BA,B be fixed Hilbert–Schmidt operators on H=ℓ2(N;F)H=\ell^2(\mathbb{N};\mathbb{F}), where F=R\mathbb{F}=\mathbb{R} or C\mathbb{C}, and let GG be a standard Gaussian array over the same field. Write s(T)=(sn(T))n≥1s(T)=(s_n(T))_{n\geq1} for the singular-value sequence of a compact operator TT, and I(T)\mathcal{I}(T) for the algebraic two-sided principal ideal that it generates. The Gaussian product satisfies

I(AGB)=I(s(A)⊙s(B))(1)\mathcal{I}(AGB)=\mathcal{I}(s(A)\mathbin{\odot}s(B)) \tag*{(1)}

almost surely. Here I(a)=I(diag⁡(an))\mathcal{I}(a)=\mathcal{I}(\operatorname{diag}(a_n)) for a nonnegative nonincreasing null sequence aa. The operation ⊙\odot is defined by multiplication of square tails: if a,ba,b are nonnegative, nonincreasing, and square summable, then

Fa(n)=∑k≥nak2,Fa⊙b(n)=Fa(n)Fb(n),(2)F_a(n)=\sum_{k\geq n}a_k^2,\qquad F_{a\odot b}(n)=F_a(n)F_b(n), \tag*{(2)}
(a⊙b)n2=an2Fb(n)+bn2Fa(n+1)=∑min⁡(i,j)=nai2bj2.(3)(a\odot b)_n^2=a_n^2F_b(n)+b_n^2F_a(n+1)=\sum_{\min(i,j)=n}a_i^2b_j^2. \tag*{(3)}

Thus, on one event for the fixed pair A,BA,B, membership of AGBAGB in every algebraic two-sided operator ideal is determined by the same profile. Principal-ideal equality permits constant factors and finite index dilations in comparing singular values; the individual singular values may vary with the outcome. The notation AGBAGB refers to a Hilbert–Schmidt-valued Gaussian series, defined in (23); it does not require the unweighted array GG to define a bounded operator.

The deterministic operation in (1) raises two questions: which numerical arrays realize it uniformly, and which families of factors retain the inclusion order of the input ideals? Theorems Theorem A and Theorem B answer these questions. Theorem Theorem C then combines their consequences with an operator-valued Wick identity: in a specified Gaussian realization, one fixed Wick composition restores the inclusion relations within the local common fibres constructed in Theorem B.

2020 Mathematics Subject Classification. Primary 47B10, 47L20; Secondary 60B20, 60H07, 06F05.

Key words and phrases. Gaussian products; principal ideals; rectangular tests; Wick products; order recovery; Boolean fibres.

Uniform spectral comparison and deterministic realization

We use the standard orthonormal basis (en)(e_n) of HH. The entries of GG are independent standard real Gaussians when F=R\mathbb{F}=\mathbb{R}. In the complex case they have the form (ξ+iη)/2(\xi+i\eta)/\sqrt{2}, where ξ,η\xi,\eta are independent standard real Gaussians. Singular values are arranged in nonincreasing order and extended by zero beyond the rank. Write B(H)B(H) for the bounded operators, ∥⋅∥\lVert\cdot\rVert for the operator norm, and

Sq={T compact:∑n≥1sn(T)q<∞},∥T∥q=(∑n≥1sn(T)q)1/q(0<q<∞).\mathcal{S}_q=\left\{T\text{ compact}:\sum_{n\geq1}s_n(T)^q<\infty\right\},\qquad\lVert T\rVert_q=\left(\sum_{n\geq1}s_n(T)^q\right)^{1/q}\qquad(0<q<\infty).

In particular, S1\mathcal{S}_1 and S2\mathcal{S}_2 are the trace-class and Hilbert–Schmidt operators; we also write ∥⋅∥HS\lVert\cdot\rVert_{\mathrm{HS}} for ∥⋅∥2\lVert\cdot\rVert_2. The inner product is linear in its first variable, and x⊗yx\otimes y denotes the operator z↦⟨z,y⟩xz\mapsto\langle z,y\rangle x. Explicitly,

I(T)={∑j=1mBjTCj:m∈N, Bj,Cj∈B(H)}.\mathcal{I}(T)=\left\{\sum_{j=1}^{m}B_jTC_j:m\in\mathbb{N},\ B_j,C_j\in B(H)\right\}.

No norm closure is taken. We put Da=diag⁡(an)D_a=\operatorname{diag}(a_n). The singular-value decomposition gives I(T)=I(s(T))\mathcal{I}(T)=\mathcal{I}(s(T)). Products, powers, inequalities, minima, and maxima of sequences are understood coordinatewise unless stated otherwise.

Let C\mathcal{C} be the set of nonnegative nonincreasing sequences in ℓ2\ell^2, and let C+\mathcal{C}_+ consist of its strictly positive members. § 2 verifies that a∘b∈Ca\mathbin{\circ} b\in\mathcal{C} for a,b∈Ca,b\in\mathcal{C}. On squared sequences this operation is convolution on the semilattice (N,min⁡)(\mathbb{N},\min), with the multiplicative square-tail transform in (2) [4 Example 2.5 and Lemma 2.8].

Uniform singular-value comparison for weighted Gaussian matrices. Corollary 4.1 gives one event of probability one, determined by GG, on which every a,b∈Ca,b\in\mathcal{C} defines a Hilbert–Schmidt matrix ZG(a,b)=(aigijbj)Z_G(a,b)=(a_i g_{ij}b_j) and, with ρ=a∘b\rho=a\mathbin{\circ} b,

sn(ZG(a,b))≤Cωρ⌈n/2⌉,s⌈n/λ⌉(ZG(a,b))≥cω,λρn(n≥1, λ>1).s_n(Z_G(a,b))\leq C_\omega\rho_{\lceil n/2\rceil},\qquad s_{\lceil n/\lambda\rceil}(Z_G(a,b))\geq c_{\omega,\lambda}\rho_n\qquad(n\geq1,\ \lambda>1).

The positive constants are independent of a,ba,b; the lower constant may depend on λ\lambda. In particular, on this same event,

DaGDb∈Sq⟺a∘b∈ℓq(a,b∈C, 0<q<∞).D_aGD_b\in\mathcal{S}_q\Longleftrightarrow a\mathbin{\circ} b\in\ell^q\qquad(a,b\in\mathcal{C},\ 0<q<\infty).

Corollary 4.2 gives the corresponding Schatten estimates and their uniform versions for leading finite rectangles. For example, if dn=n−1d_n=n^{-1}, then (d∘d)n∼2 n−3/2(d\mathbin{\circ}d)_n\sim\sqrt{2}\,n^{-3/2}, so

sn(DdGDd)≍n−3/2,DdGDd∈Sq⟺q>2/3.s_n(D_dGD_d)\asymp n^{-3/2},\qquad D_dGD_d\in\mathcal{S}_q\Longleftrightarrow q>2/3.

Here ≍\asymp denotes two-sided comparison by positive constants.

The common event concerns all diagonal weights in the fixed coordinate basis. For each fixed pair A,BA,B, Gaussian invariance under changes of singular-vector coordinates gives (1), with an exceptional set that may depend on A,BA,B. For a measurable random pair (A,B)(A,B) independent of GG, the same identity holds conditionally on the pair, by Corollary 4.1; the two factors need not be independent of one another.

The uniform realization is governed by a deterministic rectangular criterion. A rank flag is a strictly increasing sequence of integers 0=N0<N1<⋯0=N_0<N_1<\cdots. Let CN\mathcal{C}_{\mathcal{N}} consist of the a∈Ca\in\mathcal{C} that are constant on each set {Nj−1+1,…,Nj}\{N_{j-1}+1,\ldots,N_j\}, and put CN,+=CN∩C+\mathcal{C}_{\mathcal{N},+}=\mathcal{C}_{\mathcal{N}}\cap\mathcal{C}_+. For a numerical array M=(mij)M=(m_{ij}) over F\mathbb{F}, let M[m,k]M[m,k] denote its leading m×km\times k submatrix, and write ZM(a,b)=(aimijbj)Z_M(a,b)=(a_i m_{ij}b_j). This notation initially denotes a matrix; its existence as a Hilbert–Schmidt operator is part of the following statement.

Theorem A (Rectangular characterization). Let P,QP,Q be arbitrary rank flags. For a numerical array MM, the following conditions are equivalent.

(i) There exist C≥1C\geq1 and L∈NL\in\mathbb{N} such that, for every a∈CPa\in\mathcal{C}_P and b∈CQb\in\mathcal{C}_Q, the matrix ZM(a,b)Z_M(a,b) defines a Hilbert–Schmidt operator and, with ρ=a∘b\rho=a\mathbin{\circ} b,

sn(ZM(a,b))≤Cρ⌈n/L⌉,ρn≤Cs⌈n/L⌉(ZM(a,b))(n≥1).(4)s_n(Z_M(a,b))\leq C\rho_{\lceil n/L\rceil},\qquad\rho_n\leq C s_{\lceil n/L\rceil}(Z_M(a,b))\qquad(n\geq1). \tag*{(4)}

(ii) There exist K<∞K<\infty and κ>0\kappa>0 such that

∥M[Pj,Qk]∥2≤Kmax⁡(Pj,Qk),∥M[Pj,Qk]∥HS2≥κPjQk(j,k≥1).(5)\|M[P_j,Q_k]\|^2 \le K\max(P_j,Q_k),\qquad\|M[P_j,Q_k]\|_{\mathrm{HS}}^2\ge\kappa P_jQ_k\qquad(j,k\ge1). \tag*{(5)}

Under (ii), the upper comparison in (4) holds with dilation 22, and the lower comparison holds with every real dilation λ>K/κ\lambda>K/\kappa, with constants depending only on K,κ,λK,\kappa,\lambda. No upper dilation less than 22 is valid uniformly over all flag pairs and all arrays satisfying (ii). No relation between the flags or bound on their gaps is assumed.

The principal-ideal criterion permits a constant factor and a finite repetition of the singular-value sequence [3, 12]. Consequently, Theorem A implies

I(ZM(a,b))=I(a⊙b)(a∈CP, b∈CQ).\mathcal{I}(Z_M(a,b))=\mathcal{I}(a\mathbin{\odot} b)\qquad(a\in\mathcal{C}_P,\ b\in\mathcal{C}_Q).

The full-prefix tests Pj=Qj=jP_j=Q_j=j hold almost surely for the Gaussian array. Section 4 applies the criterion and proves the Gaussian statements above, including the lower comparisons for every λ>1\lambda>1.

Order determination and local common fibres

Let P2={I(a):a∈C}\mathfrak{P}_2=\{\mathcal{I}(a):a\in\mathcal{C}\}. The square-tail product induces the well-defined operation

References

  1. [1]Bingham, N. H., Goldie, C. M. and Teugels, J. L. (1987). Regular Variation. Encyclopedia of Mathematics and its Applications, vol. 27. Cambridge University Press, Cambridge. doi:10.1017/CBO9780511721434.DOI
  2. [2]Blass, Andreas (2010). Combinatorial cardinal characteristics of the continuum. In Handbook of Set Theory, M. Foreman and A. Kanamori (eds.), 395–489. Springer, Dordrecht. doi:10.1007/978-1-4020-5764-9_7.DOI
  3. [3]Calkin, John W. (1941). Two-sided ideals and congruences in the ring of bounded operators in Hilbert space. Annals of Mathematics (2) 42, no. 4, 839–873. doi:10.2307/1968771.DOI
  4. [4]Choi, Yemon (2013). Approximately multiplicative maps from weighted semilattice algebras. Journal of the Australian Mathematical Society 95, no. 1, 36–67. doi:10.1017/S1446788713000189.DOI
  5. [5]Davidson, Kenneth R. and Szarek, Stanisław J. (2001). Local operator theory, random matrices and Banach spaces. In Handbook of the Geometry of Banach Spaces, vol. 1, W. B. Johnson and J. Lindenstrauss (eds.), 317–366. North-Holland, Amsterdam. doi:10.1016/S1874-5849(01)80010-3.DOI
  6. [6]Fan, Ky (1951). Maximum properties and inequalities for the eigenvalues of completely continuous operators. Proceedings of the National Academy of Sciences of the United States of America 37, no. 11, 760–766. doi:10.1073/pnas.37.11.760.DOI
  7. [7]Gross, Leonard (1975). Logarithmic Sobolev inequalities. American Journal of Mathematics 97, no. 4, 1061–1083. doi:10.2307/2373688.DOI
  8. [8]Hinrichs, Aicke, Krieg, David, Novak, Erich, Prochno, Joscha and Ullrich, Mario (2021). Random sections of ellipsoids and the power of random information. Transactions of the American Mathematical Society 374, no. 12, 8691–8713. doi:10.1090/tran/8502.DOI
  9. [9]Hu, Yaozhong and Yan, Jia-An (2009). Wick calculus for nonlinear Gaussian functionals. Acta Mathematicae Applicatae Sinica, English Series 25, no. 3, 399–414. doi:10.1007/s10255-008-8808-0.DOI
  10. [10]Itô, Kiyosi (1951). Multiple Wiener integral. Journal of the Mathematical Society of Japan 3, no. 1, 157–169. doi:10.2969/jmsj/00310157.DOI
  11. [11]Janson, Svante (1997). Gaussian Hilbert Spaces. Cambridge Tracts in Mathematics, vol. 129. Cambridge University Press, Cambridge. doi:10.1017/CBO9780511526169.DOI
  12. [12]Kaftal, Victor and Weiss, Gary (2011). B(H) lattices, density and arithmetic mean ideals. Houston Journal of Mathematics 37, no. 1, 233–283. arXiv:0707.3115.
  13. [13]Nualart, David (2006). The Malliavin Calculus and Related Topics, 2nd ed. Probability and Its Applications. Springer, Berlin. doi:10.1007/3-540-28329-3.DOI
  14. [14]Simon, Barry (2005). Trace Ideals and Their Applications, 2nd ed. Mathematical Surveys and Monographs, vol. 120. American Mathematical Society, Providence, RI. doi:10.1090/surv/120.
  15. [15]Vershynin, Roman (2012). Introduction to the non-asymptotic analysis of random matrices. In Compressed Sensing: Theory and Applications, Y. C. Eldar and G. Kutyniok (eds.), 210–268. Cambridge University Press, Cambridge. doi:10.1017/CBO9780511794308.006.DOI

Paper details

vlnc:W_4a6fdp2n
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 1

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

GPT-6.0 (OpenAI)

View statement
Looking for related papers…
Contents