WorksMathematics
Gaussian Exterior Dilation and Schatten Classes of Lévy Area
Abstract
We study the skew-adjoint Hilbert–Schmidt operator obtained by evaluating a fixed Hilbert two-form on a centered Gaussian linear map whose covariance is the inner product on its domain multiplied by a fixed nonnegative trace-class operator. We compute the expected regularized determinant of the resulting random operator and compare the random determinant with this expectation by dilation of the scalar parameter. For each fixed covariance and two-form, a single dilation almost surely applies to every nonnegative parameter value and yields simultaneous bounds for all Schatten norms and quasi-norms of orders . The least admissible dilation satisfies a universal weak- estimate with optimal tail exponent. For Wiener area, a rank-one perturbation identity identifies the mean determinant with the regularized determinant of an explicit deterministic skew-adjoint operator and reduces Schatten membership to conditions on the covariance eigenvalues. Consequently, for each , the area belongs almost surely to the Schatten class of order if and only if the covariance of the Wiener process does. At , the criterion is finite covariance entropy; for , trace-class covariance suffices.
Introduction
Lévy area is the antisymmetric part of a second-order stochastic integral. It enters the construction of rough paths and the analysis of differential equations driven by irregular signals [7, 15]. For a Hilbert-space Wiener process with trace-class covariance, the area has a Hilbert–Schmidt realization. We determine when it belongs to the trace class and to the Schatten ideals below trace class, in terms of the covariance eigenvalues.
The main result concerns a fixed Hilbert two-form evaluated by a centered Gaussian linear map whose covariance is the inner product on its domain multiplied by a fixed nonnegative trace-class operator. The resulting skew-adjoint operator has a regularized determinant whose mean is given by an exterior-power expansion. We prove that a single random dilation compares this determinant with its mean for every nonnegative value of the scalar parameter. The least dilation satisfies a universal weak- estimate, and the tail exponent is optimal. For each fixed covariance and two-form, the comparison holds on one event of probability one and gives simultaneous upper and lower bounds for all Schatten norms and quasi-norms of orders strictly between zero and two. Here the pathwise comparison is simultaneous in the determinant parameter for a fixed Gaussian image; the Wiener-area application concerns the area over a fixed time interval.
For Wiener area, the mean determinant is the regularized determinant of an explicit deterministic skew-adjoint operator. A rank-one perturbation identity determines its determinant function and the exterior coefficients of the Brownian time kernel. It also reduces Schatten membership to scalar summability conditions. For every order , the area belongs almost surely to the Schatten class of order if and only if the covariance belongs to the same class. For a nonzero covariance with positive eigenvalues , set . The trace-class criterion is
Thus the endpoint requires a logarithmic summability condition in addition to the trace-class assumption on the covariance.
The conclusion for Schatten orders greater than one follows from the exactness criterion for tensor norms of Ledoux, Lyons and Qian [13]; the corresponding estimate is recorded after the proof of Corollary 1. Deierich [5] develops finite-dimensional approximation for Banach-space Wiener rough paths. Grong, Nilssen and Schmeding [8] study geometric approximation in infinite-dimensional spaces for a fixed Schatten tensor norm.
For general continuous Hilbert-space semimartingales, the author [18] proves almost-sure joint continuity of stochastic cross integrals in the two time variables with values in the little weak-trace ideal. This ideal consists of compact operators whose singular values, multiplied by their indices, tend to zero. That paper also constructs Wiener examples for which this convergence can be arbitrarily slow. The covariance criteria proved here use Gaussian structure to characterize membership in the trace class and in Schatten ideals of order below one.
2020 Mathematics Subject Classification. Primary 60G15, 47B10; Secondary 60H05, 15A75.
Key words and phrases. Gaussian exterior forms; Lévy area; regularized determinants; Schatten classes; covariance entropy.
Distributional formulas for stochastic area originate in Lévy’s work [14]. Higher-dimensional formulas and double Wiener–Itô representations are developed in [6, 9, 11]. Dela Cruz and Oberhauser [4] determine finite-dimensional singular-value distributions and high-dimensional asymptotics for matrices of Brownian Lévy areas. Their Section 2.4 also treats the singular-value densities of area matrices for sufficiently regular centered Gaussian processes with independent identically distributed coordinates. The present paper gives almost-sure Schatten membership criteria for a fixed Hilbert-space Wiener process with arbitrary trace-class spatial covariance.
The proof of the Gaussian comparison uses the covariance geometry of exterior contractions. Alternation bounds the largest covariance eigenvalue of a Gaussian contraction by its covariance trace divided by the exterior degree. For degree at least two, Gaussian integration by parts converts this relation into an inverse-norm estimate. Successive contractions, followed by a one-dimensional Gaussian density bound, give a normalized small-ball estimate whose constant grows linearly with the degree. Applying this estimate and Markov’s inequality at geometric thresholds makes the failure probabilities summable over the determinant coefficients. The resulting simultaneous coefficient bounds yield a single dilation for the whole determinant. Gaussian conditioning also underlies Barvinok’s estimates for determinant estimators of mixed discriminants [3 Sections 3–4]. The estimate used here exploits alternation to control all exterior degrees in the same comparison.
The deterministic operator in the Wiener-area application is the skew-adjoint part of the upper triangular truncation of a rank-one positive operator. It therefore belongs to the classical triangular-projection framework [12]; the relevant weak-type estimate is recalled in [1 Section 8]. The associated rank-one perturbation has the classical triangular form recalled in [2 Section 2.3]. We use a direct determinant calculation for this model to identify the mean determinant and obtain the covariance summability criteria.
Notation and preliminaries. Throughout, Hilbert spaces are real and separable. Their inner products and norms are denoted by and . For a bounded operator , let denote its adjoint and its operator norm; denotes the identity on the space in question, and denotes the trace. We use for the Hilbert tensor product and identify with the rank-one operator . For a compact operator between Hilbert spaces, let be its singular values, counted with multiplicity and padded with zeros if necessary. For , set
For , the quantity is a quasi-norm; for , it is a norm on . The cases and are the trace-class and Hilbert–Schmidt ideals, respectively. Probability and expectation are denoted by and , and denotes the indicator of a set .
Determinants of real operators are taken after complexification. For trace-class , let denote the Fredholm determinant. For Hilbert–Schmidt , define the regularized Fredholm determinant by
The operator is trace class, so the right-hand side is defined. We use the standard continuity and product properties of these determinants [17].
For a Hilbert space and an integer , let denote the Hilbert completion of its algebraic exterior power, normalized so that increasing wedges of an orthonormal basis are orthonormal; set . For nonnegative integers , exterior multiplication extends to a continuous bilinear map , and
This estimate follows by orthogonal antisymmetrization in . For a bounded operator between Hilbert spaces, its induced exterior power is determined by
In degree zero, is the identity on . We write for the -fold exterior product of , with . In degree two, let denote the isometric alternating embedding determined by
The assignment extends linearly and continuously to an isomorphism from onto the skew-adjoint Hilbert–Schmidt operators on . For such an operator , let denote the corresponding two-form. Then . If the nonzero eigenvalue pairs of are , with , define its determinant function, with scalar parameter , by
Indeed, for a suitable orthonormal family , the orthogonal block representation is , and each exterior coefficient is the corresponding elementary symmetric function of . The series and product define the same entire function, also denoted by , of the complex parameter.
Pathwise dilation for Gaussian exterior forms
Let be a probability space. Let be Hilbert spaces, and let be trace class on . Let be linear, with the scalar family centered and jointly Gaussian. Assume
Gaussian fourth moments imply for . The alternating bilinear map therefore induces a linear map from the algebraic exterior square of into , which we call Gaussian evaluation. No samplewise bounded operator is assumed. Write for the finite or countable sequence of positive eigenvalues of , counted with multiplicity. For integers , set
An empty sum is zero. These quantities are finite and satisfy .
Theorem 1 (Pathwise dilation). Gaussian evaluation has a unique continuous linear extension from to . For each fixed , let be its image and let be the associated skew-adjoint operator. Then , and
The series on the right defines an entire function. With , define the least pathwise dilation by
Then is a measurable -valued random variable and satisfies
with a constant independent of and . For , define the deterministic quantity by
There is an event with such that, on , and both inequalities in (4) hold with . On the same event,
with infinite values allowed, and for every $p\ge 2.
Proof. Step 1. Gaussian evaluation and its mean determinant. Let . Equation (2) implies that and . Choose an orthonormal eigenbasis of with , and put . With denoting the Kronecker delta, we have
Joint Gaussianity implies that the are independent isonormal Gaussian processes over : each is a centered Gaussian linear map from to with covariance . Moreover, in . For background on isonormal processes, see [10].
Let be an isonormal Gaussian process over . For an integer and , let be the adjoint of left exterior multiplication . For and an orthonormal basis of , define
To check convergence, expand in an orthonormal wedge basis. Each squared coefficient occurs exactly times in . The series is basis independent by the isonormal isometry applied to the Hilbert--Schmidt map . If is measurable with respect to a -field independent of , the same partial sums define in and satisfy
Conditionally on , is centered Gaussian on with covariance . This series converges in trace norm almost surely and has trace . For an integer , distinct indices $i_1,\ldots,i_r from the chosen covariance eigenbasis, and , define the successive contractions by this conditional construction. At step , the input has degree and is measurable with respect to the processes used in the preceding steps, hence independent of . Iterating the conditional second-moment identity gives
Define the Gaussian evaluation of by
The convergence follows from (8) with and the finiteness of . Regard as an element of , and let be its associated skew-adjoint operator. The same identity gives
Each coordinate depends linearly on , so (9) also applies to differences of inputs. On the algebraic exterior square, expansion of the coordinates identifies this map with the substitution . Density therefore proves that the construction is the unique continuous extension of Gaussian evaluation. In particular, it is independent of the chosen eigenbasis.
For each integer , set . For and an increasing multi-index , write and . We claim that
almost surely, simultaneously for all such pairs . For , this is the definition of . If is supported on a finite-dimensional subspace , choose representatives of the images under of a basis of . Outside a single null set, these vectors define a linear map . Its induced maps on exterior powers preserve exterior products, and determinant expansion gives (10) for every and every increasing multi-index of length . If , both sides vanish.
For general , choose finite-rank orthogonal projections increasing strongly to . For every fixed integer , convergence on simple wedges and the bound imply, by density, that converges strongly to the identity on . Set and denote its Gaussian image by . Equation (9) gives
For each fixed , continuity of exterior multiplication gives in . Thus the right-hand side of (10), with replaced by , converges in by (8). For fixed , the left-hand side is a polynomial in the finitely many coordinates of indexed by pairs of elements of . It therefore converges in probability to the corresponding coordinate of . Uniqueness of limits in probability proves (10) almost surely for this pair . Since there are only countably many such pairs, the identities hold on a single event of probability one for the fixed input .
Squaring and summing (10), then applying Tonelli, yields
Equation (1) and another application of Tonelli prove (3). Since , the product representation in (1), applied to the skew-adjoint operator associated with , gives, for ,
Here the last inequality uses . The series therefore converges uniformly on compact subsets of and defines an entire function.
Step 2. A small-ball estimate for the exterior coefficients. Fix , where , and put . The operator defined by satisfies
For , complete to an orthonormal basis of . Contraction removes this basis vector from each wedge in which it occurs, giving . The Hilbert--Schmidt identity follows from (7). The covariance of is , so
Thus the exterior degree bounds from below the ratio of the covariance trace to its operator norm. This relation gives the inverse-norm estimate below.
First suppose that is finite dimensional. For , set . Gaussian integration by parts, applied to the smooth vector field , gives
Here we used (11) and . The left-hand side is at most . Monotone convergence as therefore proves
We interpret the reciprocal of zero as . The calculation also applies to singular covariance, since integration by parts may be carried out in the underlying independent Gaussian coordinates.
For general , let , with the finite-rank projections used above. Then in , and (7) gives in . For all sufficiently large , . The finite-dimensional estimate applies to and the restriction of to . Along an almost surely convergent subsequence, Fatou’s lemma and give (12). In particular, almost surely.
Fix an integer and a nonzero vector . On an auxiliary probability space, let be independent isonormal Gaussian processes over , and put
Let be the trivial -field and let be generated by . Each is -measurable and square integrable. We prove inductively that for . This is immediate for . Suppose that and . Then almost surely and has exterior degree . Conditional on , we may therefore apply (12) with , since is independent of . This gives
Taking expectations proves the induction step. Iterating the estimates yields
In particular, almost surely. For the final contraction, set . Conditional on , is a centered real Gaussian random variable with variance . Its conditional density gives
Repeated application of (7) gives . Combining the conditional density bound with the preceding inverse-moment estimate and setting , we obtain
We next apply (13) to the coefficients of the determinant. For a fixed with positive mean coefficient in (3), both and are positive. Define
For each , (13) applies to with and , since are independent isonormal processes. The weights are deterministic, nonnegative, and sum to one. By (10), the normalized coefficient is
For , the event implies that the indices with carry at least half of the total weight. Hence
Taking expectations in the preceding inequality and using Tonelli gives
Apply (13) with and use . Also, . Consequently,
The bound at follows by continuity of probability from above. No independence between the variables is required. For every , the bounds in (14) at the thresholds are summable over . This is the lower-tail estimate used to compare all determinant coefficients by one dilation.
Step 3. The least admissible dilation. Choose a measurable representative of that is skew-adjoint and Hilbert–Schmidt at every outcome, setting it equal to zero on a null exceptional set. For each realization, let be the set of admissible dilations in (4). Since is nondecreasing, is upward closed. Continuity of and implies that is closed in and that admissibility can be tested at nonnegative rational values of . Consequently, for every finite ,
For each fixed , the function is continuous on . Every event in this countable intersection is therefore measurable, which proves the measurability of . If is nonempty, its infimum belongs to ; hence is admissible whenever it is finite.
Let be the set of indices for which the coefficient of in is positive. By nonnegativity and countability, all coefficients with zero mean vanish on a common event of probability one. On that event, the bounds
imply for every , by summing the nonnegative power series term by term. Thus, for , Markov's inequality and (14) yield
For , the right side multiplied by is at most . For , use probability at most one. This proves (5), including almost sure finiteness. If is empty, then and almost surely, so .
Step 4. Simultaneous Schatten comparison. For a skew-adjoint , we first verify the integral representation
Let be the nonzero eigenvalue pairs of . For ,
All summands are nonnegative. For and , integration by parts after the substitution gives
The boundary terms vanish because at zero and at infinity; for , both sides are zero. Tonelli's theorem and now prove (16), with both sides allowed to be infinite.
Let be the intersection of the probability-one events obtained in Steps 1 and 3. On this event, , , and the inequalities in (4) hold with for every . Fix a realization in . Then
For any , integrate these nonnegative functions against . The changes of variables in the lower integral and in the upper integral, together with (16), give
Taking -th roots proves (6). These operations remain valid for infinite integrals because on the fixed realization. Since was arbitrary and the event was chosen independently of , the comparison holds simultaneously for every . On the same event, for every by .
□
Remark (Sharpness of the tail exponent). The power in (5) is optimal. Take , let be the standard orthonormal basis of , and set and . Let be the determinant of a matrix with independent standard normal entries. Then
An orthogonal change of the four Gaussian coordinates expresses as
where are independent standard Gaussian variables. The two terms are independent exponential variables with mean one, so has density on . For , the events and are disjoint. Consequently,
In particular, and . The tail exponent in (5) is therefore optimal; the numerical constant 9 is not claimed to be optimal.
Remark (Comparison at large parameters). Suppose that has infinitely many nonzero coefficients. Then every coefficient of even degree is positive: vanishing of or forces all subsequent coefficients to vanish. Thus is defined for every . For every , Markov’s inequality and (14) give
Apply the first Borel–Cantelli lemma to a countable sequence . On one event of probability one,
Intersect this event with the event of Theorem 1. For each resulting realization, fix any and choose . There is an integer such that for every . Consequently, each of
has nonnegative coefficients in degrees at least , and its coefficient of degree is strictly positive. Its remaining terms form a polynomial of degree less than . Thus both differences are nonnegative for all sufficiently large .
The conclusion holds on the same event for every real ; only the threshold in depends on the realization and on . The infimum of the eventually admissible dilations is therefore one. The value one is eventually admissible if and only if identically: equality on an interval extends to the whole complex plane by analyticity. The least global dilation also depends on the behavior at bounded parameter values.
Wiener area and covariance entropy
Let be an -valued Wiener process with and covariance , , on a filtered probability space satisfying the usual conditions. Write for the positive eigenvalues of and . Its skew area on is
The first integral in (17) is an -valued Itô integral. Indeed, the Itô isometry gives
The second integral is the adjoint of the first, since taking adjoints is a bounded real-linear isometry on . Thus is a skew-adjoint Hilbert–Schmidt operator. The Itô–Stratonovich correction is symmetric and cancels in the antisymmetric difference, so either convention gives the same area. Standard Hilbert-space stochastic integration is treated in [16].
Corollary 1 (Wiener area). For each fixed trace-class covariance , there is a single event of probability one on which for every and, simultaneously for every ,
For , the entropy sum is defined to be zero. The mean determinant is
Choose an ordered orthonormal eigenbasis of , with , in any ordering of the positive eigenvalues. Let denote the sign function, with . Define by its matrix entries :
Extend by zero on . This gives a skew-adjoint Hilbert–Schmidt operator. Then , and the nonzero singular values of , including multiplicities, are independent of the chosen ordered eigenbasis. The dilation factor defined by (4) with satisfies (5) and, on the same probability-one event,
The deterministic operator also satisfies, with infinite values allowed,
Proof. The case is immediate; assume henceforth that .
Step 1. A rank-one determinant identity. Work on , equipped with the ordered eigenbasis fixed above. Both and vanish on , so this restriction does not change their determinants. Set
Then
The rank-one perturbation is upper triangular with diagonal entries : adding cancels each entry of below the diagonal. After complexification, has the form of Livšic’s triangular model for operators with rank-one imaginary part; see [2 Section 2.3]. The determinant identity below will determine both the exterior coefficients of the time form and the trace-class criterion. For , define
This is well defined because . We claim that
Suppose first that has finite rank. By the preceding triangularity and the matrix determinant lemma,
For real , the inverse exists because , and
The resolvent is self-adjoint and commutes with . Thus is skew-adjoint and has zero quadratic form on real vectors. In this finite-dimensional calculation, , so . These identities prove (23) when has finite rank.
Now suppose that has infinite rank. Let be the orthogonal projection onto the first covariance coordinates and put , , and . Writing and , one has
In particular, in operator norm. For real , both and are positive contractions. For every fixed , their resolvent identity gives, uniformly for ,
The ordinary Fredholm determinants converge by trace-norm continuity, and the regularized determinants by Hilbert–Schmidt continuity [17]. Together with , these convergences allow passage to the limit in the finite-rank identity, proving (23) for every real .
Both and are even, so taking even parts yields
Both sides of (24) are entire, so the identity extends to every complex parameter.
Step 2. The exterior coefficients of the time form. Let and define the integral operator by
The kernel is square integrable and antisymmetric, so is skew-adjoint and Hilbert–Schmidt. Let be the corresponding two-form. Under the identification of with , the normalized alternating embedding satisfies . For each integer and , set . Let be the orthogonal projection of onto the span of the normalized indicators , . Put , regarded as an operator on which vanishes on the orthogonal complement of this span. Its matrix is
The kernel of agrees with off the diagonal partition squares and is zero on those squares. Hence
Apply (24) in dimension with all covariance eigenvalues equal to . The matrix in that identity has entries , so it is precisely the matrix of on . Hence
locally uniformly for . Hilbert–Schmidt convergence and continuity of the regularized determinant identify the limit as . Comparing coefficients in (1) gives
Step 3. Identification with Wiener area. For each , define , . These are independent standard Brownian motions, and
For , define
The series defining converges in , since its second moment is . The map satisfies (2), and in for every .
For , define the mixed Gaussian integral first on simple tensors , with , by
Independence and the isonormal isometry imply that extends uniquely to a linear isometry
For every ,
This identity follows first for simple wedges and then by continuity. Under the tensor convention fixed above, the off-diagonal entries of (17) are
The equality follows first for step-process approximations and then by the Itô isometry. Since , these entries are precisely those of the Gaussian evaluation of in (10). The diagonal entries vanish, and both operators vanish on . Moreover,
It follows that the Gaussian evaluation of under agrees with the Wiener area in . For the double Wiener–Itô approach to Gaussian area, see also [6].
Equations (3), (25), and the even Fredholm expansion now give
This proves (19) and . By (1), the zeros of this determinant determine the nonzero singular values with their multiplicities. Since depends only on the unordered covariance spectrum, these singular values are independent of the chosen ordered eigenbasis. By (16), the identity gives for every , with infinite values allowed. Intersect the event on which the Gaussian evaluation and the Wiener area agree with the event in Theorem 1. On this common event, (6) gives (21) simultaneously for every . The same dilation factor satisfies the tail bound (5). We work on this event for the remainder of the proof.
Step 4. Schatten membership and the trace-class endpoint. It remains to determine the Schatten classes containing and to prove (22). The nonzero singular values of occur in equal pairs, so . Consequently,
For , define
The Fredholm product formula shows that both logarithms have derivative at zero. Equation (23) gives
The bounds on imply
Moreover, , since its coefficients are nonnegative and the coefficient of is . Therefore
The function is integrable against for every .
For each and , set
The Fredholm product formula and give . Each is nonnegative, since and
Fix , and let be Euler’s gamma function, for . If , then . Otherwise . Since , Tonelli’s theorem and the identity
give
The inner integral equals by Euler’s beta integral. For , use ; at , the last integral equals . We obtain
Since , Tonelli permits summation without any prior finiteness assumption:
For , this integral is finite exactly when . At , it is the covariance entropy. For , it is always finite, since . Equations (26) and (16) therefore give the corresponding membership criteria for . The comparison (21) proves (18) and for on the same event. Hilbert–Schmidt membership gives for every .
At , (16) and a change of variables give
Integrating (26) and using (28) proves (22). □
Remark (Schatten orders greater than one). The conclusion for also follows from the exactness criterion for tensor norms of Ledoux, Lyons and Qian [13 Definition 1 and Theorem 3]. Let be a Hilbert space, let be trace class on , and let , , be mutually independent centered -valued Gaussian random variables with covariance . For every integer ,
For , this follows by interpolation between the trace-norm bound and the Hilbert–Schmidt bound . For , the Hilbert–Schmidt bound suffices. In every case the exponent of is strictly less than one, as required by the exactness criterion.
For , set . The entropy in (18) then equals . Thus, for each fixed covariance, the area is trace class almost surely if and only if the normalized covariance spectrum has finite entropy. The Hilbert–Schmidt norm satisfies
If has rank one, then and . If has rank two, then .
Remark (Covariance spectra). For positive sequences and , write if there are constants such that for all sufficiently large . Write the positive covariance eigenvalues as . For each fixed covariance considered below, all membership statements hold on the single probability-one event of Corollary 1.
(i) Suppose that
Then , so the covariance entropy is finite. Moreover, exactly when . Consequently,
In particular, .
(ii) Suppose that
The covariance is trace class, while . Thus
For every , the area belongs to for all and to none of the classes with , since diverges for every .
Declaration of AI use
The author used ChatGPT (GPT-6.0; OpenAI) interactively in developing proofs and refining the presentation. The author has independently verified all proofs and takes full responsibility for the content of the manuscript.
References
School of Mathematical Sciences, University of Science and Technology of China, Hefei, Anhui 230026, China
Email address: [email protected]
References
- [1]Aleksandrov, A. B. and Peller, V. V. (2022). Triangular projection on S_p, 0 < p < 1, and related inequalities. Preprint. arXiv:2207.02975v1.arxiv.org/abs/2207.02975
- [2]Arlinskiĭ, Y. and Tsekanovskiĭ, E. (2006). Non-self-adjoint Jacobi matrices with a rank-one imaginary part. J. Funct. Anal. 241, no. 2, 383–438. doi:10.1016/j.jfa.2006.05.002. Author preprint: arXiv:math/0602033v1.DOI
- [3]Barvinok, A. (1999). Polynomial time algorithms to approximate permanents and mixed discriminants within a simply exponential factor. Random Structures Algorithms 14, no. 1, 29–61. doi:10.1002/(SICI)1098-2418(1999010)14:1<29::AID-RSA2>3.0.CO;2-X.doi.org/10.1002/(sici)1098-2418(1999010)14:1<29::aid-rsa2>3.0.co;2-x
- [4]Dela Cruz, D. J. and Oberhauser, H. (2026). The singular values of Lévy’s area matrix. Preprint. arXiv:2606.08928v1.arxiv.org/abs/2606.08928
- [5]Dereich, S. (2010). Rough paths analysis of general Banach space-valued Wiener processes. J. Funct. Anal. 258, no. 9, 2910–2936. doi:10.1016/j.jfa.2010.01.018. Author manuscript.DOI
- [6]Ferreiro-Castilla, A. and Utzet, F. (2011). Lévy area for Gaussian processes: A double Wiener–Itô integral approach. Statist. Probab. Lett. 81, no. 9, 1380–1391. doi:10.1016/j.spl.2011.04.015. Author preprint: arXiv:1007.2516v1 (2010).
- [7]Friz, P. K. and Hairer, M. (2020). A Course on Rough Paths: With an Introduction to Regularity Structures, 2nd ed., Universitext, Springer, Cham. doi:10.1007/978-3-030-41556-3.DOI
- [8]Grong, E., Nilssen, T. and Schmеding, A. (2022). Geometric rough paths on infinite dimensional spaces. J. Differential Equations 340, 151–178. doi:10.1016/j.jde.2022.08.034. Author version: arXiv:2006.06362v4 (2021).DOI
- [9]Helmes, K. and Schwane, A. (1983). Lévy’s stochastic area formula in higher dimensions. J. Funct. Anal. 54, no. 2, 177–192. doi:10.1016/0022-1236(83)90053-8.DOI
- [10]Janson, S. (1997). Gaussian Hilbert Spaces, Cambridge Tracts in Mathematics, vol. 129, Cambridge University Press, Cambridge. doi:10.1017/CBO9780511526169.DOI
- [11]Krée, P. (1986). A remark on Paul Lévy’s stochastic area formula. In Aspects of Mathematics and its Applications (Barroso, J. A., ed.), North-Holland Mathematical Library, vol. 34, North-Holland, Amsterdam, pp. 533–550. doi:10.1016/S0924-6509(09)70279-X.DOI
- [12]Kwapień, S. and Pełczyński, A. (1970). The main triangle projection in matrix spaces and its applications. Studia Math. 34, no. 1, 43–67. doi:10.4064/sm-34-1-43-67.DOI
- [13]Ledoux, M., Lyons, T. and Qian, Z. (2002). Lévy area of Wiener processes in Banach spaces. Ann. Probab. 30, no. 2, 546–578. doi:10.1214/aop/1023481002.DOI
- [14]Lévy, P. (1951). Wiener’s random function, and other Laplacian random functions. In Proceedings of the Second Berkeley Symposium on Mathematical Statistics and Probability (Neyman, J., ed.), University of California Press, Berkeley, pp. 171–187. Project Euclid.DOI
- [15]Lyons, T. J. (1998). Differential equations driven by rough signals. Rev. Mat. Iberoam. 14, no. 2, 215–310. doi:10.4171/RMI/240.DOI
- [16]Métivier, M. (1982). Semimartingales: A Course on Stochastic Processes, De Gruyter Studies in Mathematics, vol. 2, Walter de Gruyter, Berlin–New York. doi:10.1515/9783110845563.DOI
- [17]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.
- [18]Zhao, G. (2026). Sharp weak-trace regularity of stochastic cross integrals. Preprint, version 6, 14 October 2026. arXiv:2608.08756v6.arxiv.org/abs/2608.08756