WorksMathematics
Gaussian Matrices and Exterior Forms: Deterministic Principal Ideals
Abstract
For an infinite Gaussian matrix with independent entries and doubly monotone variances, we identify its algebraic principal ideal by a deterministic shell profile. The exact boundedness condition also implies compactness. A lower singular-value estimate uniform in the mean gives an exact sum law under each fixed compact perturbation. Conditional products and spatial compression yield the corresponding classification for Gaussian evaluations of Hilbert two-forms. For Wiener area of covariance rank at least two, the generator is the covariance tail divided by its index. The area and covariance generate the same ideal precisely under regularity at infinity.
Introduction
We identify the principal ideals of Gaussian matrices with doubly monotone variances by their diagonal shell masses. Each identity below determines all algebraic two-sided ideal membership tests on one probability-one event for fixed input data and a fixed perturbation. Conditional iteration gives the ideals of Gaussian exterior evaluations; Wiener area follows by inserting its time spectrum. Write for the two-sided ideal generated by a compact operator , with no norm closure, and for a decreasing nonnegative null sequence . Equality of principal ideals means mutual domination of singular-value sequences up to constants and finite index dilations.
Let , where the are independent standard real Gaussian variables and , . A bounded realization exists almost surely exactly when the first row and first column of are summable. It is then compact, and Theorem A gives, for every fixed deterministic compact ,
Here is the variance in the th diagonal shell. The sum is generated by : a prescribed perturbation cannot cancel either ideal contribution. The classification includes nonseparable variances and compact realizations outside the Hilbert--Schmidt class.
For separable variances , the tail sums of are products of those of and . Conditioning allows this rule to be iterated. Let be trace class, let be skew-adjoint and Hilbert--Schmidt, and write , , with zeros appended in finite rank. The Gaussian evaluation is defined in , where ; no samplewise bounded map is required. For , , and , Theorem 3.1 gives
The square tail records the factors , , . The canonical form of gives the upper ideal inclusion; an independent odd--even covariance block gives the reverse. Rank at most one gives zero alternating evaluation. A deterministic tail estimate then shows that for every covariance of rank at least two precisely when for some finite , the regularity-at-infinity condition for .
2020 Mathematics Subject Classification. Primary 47B10, 60G15; Secondary 60B20, 60H05, 15A75.
Key words and phrases. Gaussian matrices, principal ideals, monotone variances, Gaussian exterior forms, regularity at infinity, Lévy area.
The Wiener time operator has and satisfies this condition. Hence
Moreover, , with equality exactly when . For example, if , then , so almost surely.
Related work
Calkin’s correspondence [2] describes operator ideals by singular-value sequences; principal ideals and finite dilations are treated in [8], and Schatten classes in [12]. The tail-average criterion used here belongs to the arithmetic-mean theory and regularity at infinity of Kaftal and Weiss [7]. Our use of that criterion concerns the generators produced by Gaussian evaluation.
Bandeira and van Handel [1] prove sharp nonasymptotic operator-norm estimates for matrices with independent entries. Latała, van Handel and Youssef [9] give dimension-free Schatten-norm, moment and tail comparisons for , including a characterization of bounded infinite Gaussian matrices and nonsymmetric extensions. Their variance profiles are more general; our boundedness criterion specializes that theory. We identify the whole singular-value sequence up to finite dilation and determine its principal ideal under fixed compact perturbations. Rudelson and Zeitouni [11] obtain lower singular-value estimates under broad-connectedness assumptions. Here monotonicity controls the covariance spectrum after projection and gives lower estimates uniform in the deterministic mean.
For stochastic area, Ledoux, Lyons and Qian [10] develop Wiener rough paths in Banach spaces under tensor-norm exactness assumptions. Ferreiro-Castilla and Utzet [5] construct Gaussian area through double Wiener–Itô integrals and study its characteristic function. Dela Cruz and Oberhauser [4] study singular-value distributions of finite-dimensional Brownian-area matrices and their high-dimensional asymptotics. We fix the spatial covariance and time interval and determine the resulting compact operator’s principal ideal.
Gaussian exterior dilation [15] compares a random regularized determinant with its expectation through one random dilation, uniformly over nonnegative parameters, and proves a universal weak- bound with optimal tail exponent. Its Wiener-area application includes the Schatten membership criteria and the covariance-entropy condition at trace class. Here we determine principal-ideal generators, their sums under fixed compact perturbations, and ideal-preservation conditions.
Proof and organization
Gaussian comparison and concentration control tail-corner norms, giving upper singular-value bounds after deleting finitely many rows and columns. For the lower bounds, covariance eigenvalue interlacing and a Gaussian Laplace estimate control the distances of conditioned rows from the spans of the other rows, uniformly in the means. The inverse Gram identity converts these distances into spectral bounds.
Section contains the matrix theorem, its proof, and the Gaussian-product consequences. Sections and treat exterior evaluation and Wiener area. Covariance comparison is proved in Appendix A.
Gaussian matrices
All Hilbert spaces and Gaussian variables are real. We use with its standard basis , and write and for the bounded and compact operators, with the operator norm. The singular values of a compact operator are listed in nonincreasing order, with multiplicities and zero completion in finite rank. They satisfy
Operators between different separable Hilbert spaces are compared after embedding the spaces in and extending by zero. This and complexification preserve the nonzero singular values and their multiplicities. The Hilbert–Schmidt class is , with norm . We use and the standard inequalities [12]
In particular, .
For a nonnegative nonincreasing null sequence , write and , where
No closure is taken. All nonzero finite-rank operators generate the finite-rank ideal. Throughout, a two-sided ideal is closed under finite sums and multiplication on either side by bounded operators, but need not be topologically closed.
Lemma 2.1 (Principal ideals and finite dilations). For nonnegative nonincreasing null sequences ,
Moreover, for every compact , and
where the maximum is coordinatewise.
Proof. These are the principal-ideal consequences of Calkin’s correspondence [2, 8]. For completeness, a sum of bounded left–right multiples of has by (5). Conversely, the stated domination expresses as the sum of coordinate copies of multiplied by bounded diagonal quotients; the quotients are zero where . A singular-value decomposition gives the assertion for , using partial isometries on its positive singular subspaces. Finally, with the coordinate projection onto , one has , proving (7). □
A table of finite nonnegative variances is doubly monotone if
Let for independent standard Gaussian , and define
The total variance may be infinite. Initially is only an array; a bounded realization is a bounded operator with these matrix entries.
Theorem A (Spectral classification by the shell profile). Suppose that is doubly monotone. The matrix has a bounded realization almost surely if and only if
Under this condition is a nonincreasing null sequence, and the realization is unique and compact almost surely. Moreover:
For each fixed satisfying (10), one event of probability one supports, for every and , finite positive constants and such that
In particular,
For every fixed deterministic compact operator , almost surely
On an event that may depend on , the lower comparison in (11) also holds for , simultaneously for all .
If (10) fails, almost surely no bounded operator has the specified matrix entries.
An event may depend on the fixed variance table and perturbation; a prescribed countable family of perturbations can share one event. Uniformity over all compact perturbations is generally impossible, as the sample-dependent choice shows. The case is trivial, and no assertion is made at or .
For each fixed compact , almost surely if and only if ; otherwise . On this same event, simultaneously for every two-sided ideal ,
Remark (A nonseparable example). For , ,
Thus
The operator is compact but not Hilbert–Schmidt when ; for it has no bounded realization.
Remark (The upper index dilation). For , one has for , while reversing the first columns makes the nonzero block triangular with nonzero diagonal almost surely. Hence , ruling out every upper dilation . This does not settle .
Proof of the matrix theorem
Construction and upper bound. We use Gaussian comparison for ordered increment variances and the concentration inequality
for a -Lipschitz function of a finite standard Gaussian vector; see [14 Theorems 6.8 and 3.25].
Lemma 2.2. For a finite rectangular matrix , where the are independent standard Gaussian variables, put and . Then
Proof. For unit vectors , the process satisfies
Compare it with , with independent standard Gaussian . Its supremum is a sum of two Euclidean norms; Jensen’s inequality proves the bound. Comparison on the spheres follows by finite dense approximations. □
Let project onto the first coordinates, , and . Assume (10). For the leading finite sections , Lemma 2.2 bounds their increasing norms by . Hence almost surely. The associated bilinear form on finitely supported vectors is bounded and extends uniquely, giving a bounded operator with the prescribed entries. Increasing the finite corners and using monotone convergence yields
Double monotonicity makes nonincreasing, and
The corner norms decrease with , so (18) and Fatou’s lemma imply almost surely. Since
is compact. Consequently in norm. The norm-Cauchy event of these finite sections is Borel; taking their limit there and zero elsewhere defines a Borel -valued random variable. The comparison events in (6) are Borel as well, because singular values are continuous and , may be restricted to integers.
If (10) fails, the first row or column has an infinite variance sum. For with ,
since . Thus that row or column is almost surely not in , excluding a bounded realization.
A finite corner starting at is -Lipschitz in its Gaussian coordinates. For ,
Concentration, (18), and increasing finite corners therefore give, with ,
A zero makes the corner zero. Fix and put , . Then by (19), and for large . Borel–Cantelli gives the eventual upper comparison. If , then , so ; the remaining initial ratios are absorbed in the constant. Intersect over rational and use for to obtain all real on one event.
Conditional distances and fixed perturbations. For a Hilbert-space Gaussian vector with mean , its covariance is , a positive trace-class operator. The order is the usual quadratic-form order.
Lemma 2.3 (A Gaussian Laplace bound). For such a vector and ,
In particular, if , then
Here is the Fredholm determinant.
Proof. Diagonalize and multiply the one-dimensional Gaussian integrals; bounded convergence handles the infinite product, including deterministic coordinates in . The second claim follows from the min–max inequality and the mean factor being at most one. □
Lemma 2.4 (Gaussian distances and covariance eigenvalues). Let be jointly Gaussian Hilbert-space vectors with arbitrary deterministic means and joint covariance , where is trace class. Write for its decreasing eigenvalues, completed by zeros, and set . For every , almost surely
Proof. For any orthogonal projection of codimension at most , the min–max principle gives : when the right side is positive, intersect the first eigendirections with ; otherwise the inequality is immediate. Thus .
Realize the joint law as , where , the are independent centered Gaussian vectors with covariance , and the independent centered Gaussian vector has covariance . Now let project off the other rows’ span; it has codimension at most and is measurable by Gram–Schmidt with zero residuals discarded. Since is independent of the other rows and jointly, conditioning on them makes Gaussian with a fixed mean and covariance . The preceding determinant bound, Lemma 2.3, and the tower property give the first inequality. Markov gives the second, first with each positive rational , then with the infimum over all by continuity. This includes zero and finite-rank .
Lemma 2.5 (The inverse Gram identity). Let have rows . If the preceding row distances satisfy , then
Proof. The rows are linearly independent. The Schur complement formula gives . Taking traces gives the negative second moment identity [13 Lemma A.4] and the bound
The last terms are at least .
Proposition 2.6 (A lower tail bound uniform in the mean). Assume (2.5) and (2.7). For every fixed compact and integers ,
Proof. If , then and the event is empty. Otherwise set
For , the first rows of are independent Hilbert-space Gaussian vectors with fixed means and covariances , so their joint covariance dominates . Their coordinate series converge in since . The eigenvalues of are in their given order. Since for and for ,
using . Lemma 2.4 with and gives
The union bound, Lemma 2.5, and imply
Transposition replaces by . Choose the larger one; it is at least . Taking and using , , proves (27).
Completion of the proof of Theorem A. Fix and , and put . For , Proposition 2.6 gives
Borel–Cantelli proves the eventual lower comparison. When , every variance in the leading block is positive. Its determinant after adding the fixed block of is a nonzero polynomial in independent continuous variables, and is nonzero almost surely. Thus whenever , simultaneously for all , so the finite initial ratios can be absorbed in a positive constant. Intersect over rational ; for real choose rational and use .
For , combine the upper and lower events. Taking and in (2.8) proves by (2.3). For fixed , the lower bound gives . On the same event , so . This proves one inclusion in (2.10); the other follows from . (7) completes the proof.
Gaussian products
The separable case of Theorem A applies conditionally to independent Gaussian products. Their spectral profiles compose by multiplication of square tails.
For nonnegative nonincreasing , put and define
This is the squared energy of in the layer . Expanding it as proves that is nonnegative and nonincreasing. Summation gives
Since square tails determine the sequence, is commutative and associative.
The operation also respects equality of principal ideals. Put for an integer . For , one has , and similarly for . Hence
which gives
If , monotonicity of the defining formula and give . The principal-ideal criterion therefore allows either input to be replaced by any sequence generating the same ideal. Thus a deterministic generator for the ideal of an intermediate random factor can be used in the next conditional application.
Corollary 2.7 (Gaussian products). Let have independent standard real Gaussian entries, and let be fixed Hilbert–Schmidt operators. The series
converges in and almost surely along square partial sums, with
Put , , and , so that
For every fixed compact with the same source and target as ,
In particular, the matrix generates . The construction and ideal identity remain valid conditionally for random in jointly independent of , with calculated from their singular values. No moment assumption on these inputs is needed for the almost sure assertions. Finite singular-value lists are completed by zeros.
Proof. Orthogonality of the Gaussian coefficients gives
Thus the square partial sums form an -bounded -valued martingale. Hilbert-space martingale convergence proves the asserted convergence and (33). The symbol denotes this series construction; it need not be a samplewise bounded operator.
Write the positive-singular-value decompositions and . The -defined variables
are independent standard Gaussians, because their coefficient vectors are orthonormal in . For the singular-value truncations , , . This identity passes to the limit by the Gaussian isometry and
Accordingly, and , where and are the coordinate partial isometries for . Zero extensions and unused Gaussian variables handle finite singular-value lists. The variance table is doubly monotone with shell profile , so Theorem A gives . Moreover,
Applying that theorem with the fixed compact perturbation gives . Together with and the opposite inclusion obtained by addition, this proves (35).
For the conditional assertion, finite partial sums are jointly Borel in and the countable Gaussian coordinates. Convergence in is a Borel condition; set the limit to zero when it fails. Each coordinate of is Borel, since singular values are continuous and its tail sums are limits of finite sums. The ideal equality is Borel by (6). Independence and Fubini therefore apply the fixed-input result to the joint law of and . The assertion is conditional; unconditional convergence follows if . □
Lemma 2.8 (Independent bilateral Gaussian contraction). Let be nonnegative nonincreasing sequences in , let be fixed and Hilbert–Schmidt, and let be independent arrays of independent standard real Gaussians. The finite-dimensional products defining
converge in as the weights are truncated and is approximated in Hilbert–Schmidt norm. Their limit satisfies
For every fixed compact with the same source and target as ,
The event may depend on the fixed inputs and on ; a prescribed countable family of perturbations can be treated on one event. Zero and zero-completed finite-rank factors are allowed.
Proof. Set . Corollary 2.7 gives
Since is independent of , the conditional assertion defines and gives
Taking expectations proves (36). Compatibility now gives ; associativity proves (37). The same isometries applied to differences in prove continuity in , identifying this construction with the stated limit of finite-dimensional products. A countable intersection handles the prescribed family of perturbations. □
Gaussian evaluation of a Hilbert two-form
A Hilbert two-form on a real separable Hilbert space is identified with a skew-adjoint Hilbert–Schmidt operator by . With increasing orthonormal wedges normalized to have norm .
Let be trace class on a real separable Hilbert space , with positive eigenvalues in decreasing order and eigenvectors . Complete a finite eigenvalue list by zeros and put . For independent isonormal Gaussian processes over [6], define
Then . Evaluations act on and are extended by zero on ; finite-dimensional spaces are embedded in when their principal ideals are compared.
For finite-rank skew-adjoint , define by sending to . This is representation-independent by linear algebra on the finite-dimensional subspace involved; the notation does not require a samplewise bounded map .
Theorem 3.1 (Classification of Gaussian exterior evaluation). For fixed as above, finite-rank evaluation extends uniquely and continuously to every skew-adjoint in , with
Set
and
If rank , then for every fixed compact operator on , without a symmetry or Hilbert–Schmidt assumption,
In particular, almost surely. For fixed , one event accommodates any prescribed countable family of deterministic compact . If rank or , then almost surely and for every compact $D.
Thus
is the squared energy of on the shell .
Proof. For finite-rank skew-adjoint , the diagonal entries of its evaluation vanish, while independence gives variance for entry when . Summation proves (39); applying the same identity to differences gives the continuous extension and the zero cases. Henceforth assume and , and fix .
For the upper inclusion, use the canonical form
with mutually orthonormal , Hilbert–Schmidt convergence, and . The arrays and are independent standard Gaussian arrays, and the evaluation and product isometries give
Zero padding handles finite coordinate lists. Lemma 2.8 and compatibility of with principal ideals yield almost surely. Closure under adjoints and finite sums therefore gives .
For the lower inclusion, separate the even and odd covariance coordinates:
The disjoint row families are independent, and the two isometries extend this identity from finite-rank . Both weight sequences retain , since and
For , use ; the constant handles . This also covers finite support. For the same compact block of , (37) and the compatibility of give
Since , we also have . This proves ; the reverse inclusion follows from . The maximum rule for principal-ideal sums and a countable intersection finish the proof.
Rank one must be excluded: although when . For rank and , the profile shows that is almost surely nonzero finite rank if either or has finite rank, and infinite rank otherwise. The event in (42) cannot cover all compact perturbations, since the sample-dependent choice would contradict the identity.
Tail averages and regularity at infinity
For a nonnegative nonincreasing sequence , put
The tail average decreases to zero but need not be summable. A nonzero sequence is regular at infinity if
for some . For infinite support this is the classical condition in [7 Definition 4.11 and Theorem 4.12]: replacing the strict tail there by our inclusive tail adds to the average and does not change the condition. We also include nonzero finite-support sequences; the inclusive convention preserves their last nonzero entry.
Lemma 3.2 (Removing dilation from a tail estimate). For every nonnegative nonincreasing ,
Equality holds if and only if (48) holds. In particular, a finite dilation does not weaken that tail condition: for some and , the estimate
already implies (48). Zero and finite-support sequences are included.
Proof. For , , which proves (49) by the finite-dilation criterion. Equality of the ideals gives (50); evaluating it at yields
This proves the tail condition, whose converse is immediate. The first sum is empty if , and no sequence entry is divided out. For support in , ; the zero case is immediate.
Universal simplification of the exterior profile
The profile separates according to changes in the covariance and form tails:
Put and , both nonnegative, nonincreasing and square summable. The first term is at least , while gives ; the same formula gives . Hence
and for rank ,
These two ideals also replace in (42). The next equivalence makes the tail condition (54) necessary and sufficient for the first ideal to be absorbed by the second for every fixed of rank at least two. The prescribed test already detects the same condition.
Corollary 3.3 (Universal collapse of the exterior profile). Fix a skew-adjoint , with and . The following are equivalent:
(i) There is such that
:::`
(ii) For every fixed trace-class covariance of rank at least two,
(iii) Identity (55) holds almost surely for on .
For , condition (i) is regularity at infinity of . The zero and finite-rank cases are included. The event in (ii) may depend on the fixed .
Proof. Assume (i) with . For , , and therefore
Together with (51), this gives
The ideal criterion and Theorem 3.1 prove (ii), which immediately implies (iii).
Conversely, take , so . Condition (iii) and Theorem 3.1 give ; hence for some , and . Since ,
Thus . Lemma 3.2, applied to , removes the dilation and proves (i), including zero and finite support. □
Under these conditions, (42) becomes
for every fixed of rank at least two and compact . The condition holds when with , since ; this includes the Wiener time operator below.
Remark. (One form and two covariance regimes.) Take the canonical coefficients in (43). They are square summable, and
For , one has and
Thus for every fixed , and : the first contribution in (53) cannot be omitted.
For the same but , and
Now , so despite failure of (54). This covariance is itself regular at infinity, since . Thus regularity of a test covariance alone does not make it a valid substitute for the prescribed in Corollary 3.3.
Wiener area as covariance tail averaging
Let be a Wiener process in a real separable Hilbert space with trace-class covariance . Its skew area is
The Hilbert–Schmidt-valued Itô integral is well defined, since and the other integral is its adjoint; see [3 Chapter 4]. Write for the covariance eigenvalues, padded by zeros in finite rank, and put
The time kernel reduces the exterior-form classification to this tail average.
Corollary 4.1 (Wiener area and the covariance ideal). Suppose rank . For every fixed deterministic compact operator , without a skew-adjointness assumption,
In particular, on one event of probability one,
and
Thus regularity at infinity is exactly the condition that covariance tail averaging preserves the principal ideal. It includes finite-rank covariances of rank at least two. If rank , then almost surely and .
Proof. Write in an eigenbasis of . The independent isonormal processes act on . For the skew-adjoint Hilbert–Schmidt operator
the off-diagonal entries of are
Thus is the Gaussian skew compression of : the identity holds first for step kernels and then by the Itô isometry, with the mixed integral defined by the isometric extension of for .
On the complexification, the functions , , form an orthonormal basis, and direct integration gives, for ,
Consequently the full singular-value sequence, with multiplicities, is
Thus is regular at infinity, and Corollary 3.3 gives and its fixed-perturbation formula (57). Lemma 3.2, applied to , now gives and (59), including finite rank. If rank , the path lies in a fixed line and (56) vanishes. □
Brownian scaling gives the same formulas for every fixed interval , , with the path recentered at ; a common event may be chosen for any prescribed countable family of intervals and fixed compact perturbations.
For example, let with . The integral test gives , hence and
Comparison of full Gaussian covariance forms
The matrix classification is stable under two-sided comparison of the full Gaussian covariance form. The proof transfers the upper and mean-uniform lower estimates from Subsection 2.1.
Corollary A.1 (Full covariance comparability). Let satisfy (8), and let be a centered jointly Gaussian array. Suppose there are such that, for every finitely supported real coefficient table ,
Then has a bounded realization almost surely if and only if (10) holds. If that condition fails, almost surely there is no bounded realization. If it holds, the realization is unique and compact almost surely, , and all comparisons in (11) hold with in place of on one event. For every fixed compact ,
The simultaneous ideal tests in (14) hold as well. The same assertions concerning hold for a compact random independent of , without a moment assumption on $B.
Proof. Assume (10). We compare covariance forms on finite coefficient arrays, so no finite total variance is required. For each finite corner starting at , Gaussian comparison with gives expected norm at most . Its norm is -Lipschitz in standard Gaussian coordinates, since it is the supremum of unit bilinear forms whose variances are at most . Thus the finite-compression construction and (21) carry over with replaced by , giving a unique compact Borel realization and the upper comparisons for every .
Fix a compact . For , the first rows of are jointly Gaussian vectors in : their total centered second moment is at most . Their covariance dominates , with as in (28), by (A.1) and double monotonicity. After division by , Lemma 2.4 therefore gives the same row-distance bound used in Proposition 2.6. Its inverse Gram and transposition argument yields, uniformly in ,
including the trivial case .
The completion argument for Theorem A now applies. Indeed, if , the covariance of the leading block dominates times the identity, so its determinant after addition of the fixed block of is nonzero almost surely. This handles the initial lower indices; Borel–Cantelli and the same rational-parameter intersections give, for each fixed , one event supporting the upper comparisons for and the lower comparisons for for every and . For the comparisons give ; for general , the lower inclusion and prove (A.2); the ideal tests follow algebraically.
If (10) fails, the first row or column has infinite variance sum. In the row case, write . Its first coordinates have covariance at least , so Lemma 2.3 gives
Bounded convergence implies almost surely, which excludes a bounded realization. The column case is identical.
Finally, under (10), the assertions for a fixed define a Borel event in : singular values are continuous, and the finite-dilation criterion (6) and the comparisons use only countably many integer constants and rational parameters. Fubini’s theorem therefore gives the same conclusions for an independent compact random , without a moment assumption. ∎
Covariance information alone cannot replace joint Gaussianity. Let be strictly positive nonincreasing sequences in , and let be independent standard Gaussian sequences. The array is almost surely rank one, yet
It has the same full covariance form as the independent-entry Gaussian matrix with variances , whose shell profile is strictly positive at every index and whose principal ideal therefore has an infinite-rank generator.
Declaration of AI Use
The author used GPT-6.0 (OpenAI) interactively in developing and refining the mathematical arguments and the presentation of this paper. The author takes full responsibility for the correctness and mathematical content of the manuscript.
References
School of Mathematical Sciences, University of Science and Technology of China, Hefei 230026, China
Email address: [email protected]
References
- [1]Bandeira, A. S. and van Handel, R. (2016). Sharp nonasymptotic bounds on the norm of random matrices with independent entries. Annals of Probability 44, no. 4, 2479–2506. doi:10.1214/15-AOP1025.
- [2]Calkin, J. 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
- [3]Da Prato, G. and Zabczyk, J. (2014). Stochastic Equations in Infinite Dimensions, 2nd ed. Encyclopedia of Mathematics and its Applications, vol. 152. Cambridge University Press, Cambridge. doi:10.1017/CBO9781107295513.DOI
- [4]Dela Cruz, Danilo Jr and Oberhauser, H. (2026). The singular values of Lévy’s area matrix. Preprint, arXiv:2606.08928. https://arxiv.org/abs/2606.08928.
- [5]Ferreiro-Castilla, A. and Utzet, F. (2011). Lévy area for Gaussian processes: A double Wiener–Itô integral approach. Statistics & Probability Letters 81, no. 9, 1380–1391. doi:10.1016/j.spl.2011.04.015.
- [6]Janson, S. (1997). Gaussian Hilbert Spaces. Cambridge Tracts in Mathematics, vol. 129. Cambridge University Press, Cambridge. doi:10.1017/CBO9780511526169.DOI
- [7]Kaftal, V. and Weiss, G. (2010). Traces on operator ideals and arithmetic means. Journal of Operator Theory 63, no. 1, 3–46. https://jot.theta.ro/jot/archive/2010-063-001/2010-063-001-001.pdf.
- [8]Kaftal, V. and Weiss, G. (2011). B(H) lattices, density and arithmetic mean ideals. Houston Journal of Mathematics 37, no. 1, 233–283. https://www.math.uh.edu/~hjm/Vol37-1.html. Author preprint: arXiv:0707.3115.
- [9]Latała, R., van Handel, R. and Youssef, P. (2018). The dimension-free structure of nonhomogeneous random matrices. Inventiones Mathematicae 214, no. 3, 1031–1080. doi:10.1007/s00222-018-0817-x.
- [10]Ledoux, M., Lyons, T. and Qian, Z. (2002). Lévy area of Wiener processes in Banach spaces. Annals of Probability 30, no. 2, 546–578. doi:10.1214/aop/1023481002.DOI
- [11]Rudelson, M. and Zeitouni, O. (2016). Singular values of Gaussian matrices and permanent estimators. Random Structures & Algorithms 48, no. 1, 183–212. doi:10.1002/rsa.20564.
- [12]Simon, B. (2005). Trace Ideals and Their Applications, 2nd ed. Mathematical Surveys and Monographs, vol. 120. American Mathematical Society, Providence, RI. doi:10.1090/surv/120.
- [13]Tao, T. and Vu, V. (2010). Random matrices: Universality of ESDs and the circular law. With an appendix by M. Krishnapur. Annals of Probability 38, no. 5, 2023–2065. doi:10.1214/10-AOP534.DOI
- [14]van Handel, R. (2016). Probability in High Dimension. APC 550 Lecture Notes, Princeton University, version December 21, 2016. https://web.math.princeton.edu/~rvan/APC550.pdf.
- [15]Zhao, G. (2026). Gaussian exterior dilation and Schatten classes of Lévy area. Preprint, arXiv:2610.34984. https://arxiv.org/abs/2610.34984.