WorksMathematics
Gelfand numbers of Schatten class embeddings in the quasi-Banach range
Abstract
We prove Conjecture 1.4 of Prochno and Strzelecki [J. Approx. Theory 277 (2022), 105736] on Gelfand numbers of embeddings between finite-dimensional Schatten classes. For and indices bounded away from the ambient dimension by a fixed proportion, we obtain the conjectured lower bound, uniformly in the target exponent . Together with the known upper estimate obtained by Hinrichs, Prochno and Vybíral [Math. Ann., 380(3-4):1563-1593, 2021], this gives the exact asymptotic order in the previously unresolved range. As a consequence, we obtain a complete all-index description by combining our result with the estimates of Hinrichs, Prochno and Vybíral and of Prochno and Strzelecki. The proof is deterministic and is based on Hilbert–Schmidt projections on rectangular matrix spaces.
2020 Mathematics Subject Classification: Primary 47B06, 47B10; Secondary 46B20
Introduction & Main Results
The Schatten classes form one of the classical families of operator ideals in functional analysis. For a Hilbert space and , the Schatten -class consists of all compact operators on whose sequence of singular values belongs to . We write for the compact operators equipped with the operator norm. The cases and give the trace-class, or nuclear, operators and the Hilbert–Schmidt operators, respectively. The origins of the theory lie in Schatten’s work on cross-spaces of linear transformations [26] and his subsequent papers with von Neumann [28, 29]. These investigations formed part of the development of nuclear operators and tensor products, with related contributions beyond the Hilbert space setting by Ruston [25] and Grothendieck [9]. Schatten’s monograph [27] subsequently provided a systematic account of norm ideals in terms of symmetric gauge functions and unitarily invariant crossnorms. As he observed there, the classical study of individual compact operators had only comparatively recently been complemented by an interest in the structure of spaces of such operators.
This shift from operators to operator spaces naturally leads to questions about their geometry. For finite , the Schatten class is a non-commutative counterpart of , sharing its nested scale for , uniform convexity for , and, in this range, duality with the conjugate exponent under the trace pairing; see [27, 31]. Nevertheless, the geometry of Schatten classes is not determined by the corresponding sequence-space results. Building on the results of Kwapień and Pełczyński [21], Gordon and Lewis [8] proved that, on an infinite-dimensional Hilbert space, the Banach spaces , , fail to have local unconditional structure whenever . In finite dimensions, the geometry of Schatten unit balls has been investigated through isotropic constants [18], concentration and thin-shell phenomena [11, 23], asymptotic volumes and volume ratios [15], and volumes of intersections [16]. A complementary probabilistic perspective is provided by Sanov-type large deviation principles for empirical eigenvalue and singular-value distributions of random matrices [17].
For approximation theory, a particularly relevant aspect of this geometry is the behaviour of the identity mappings on subspaces of prescribed codimension. This is quantified by their Gelfand numbers. Geometrically, these numbers measure how small a section of the domain unit ball can be made in the target norm. Their determination therefore requires understanding the singular-value structure of matrices in subspaces of prescribed codimension and how it governs the comparison of different Schatten quasi-norms. This brings the distinction between commutative and non-commutative geometry directly into the problem: sharp estimates for the embeddings do not, by themselves, settle the corresponding questions for matrices.
The geometry of low-codimensional matrix subspaces that underlies the study of Gelfand numbers also plays a central role in low-rank matrix recovery and completion, where Schatten norms and quasi-norms play a role analogous to that of sequence norms and quasi-norms in sparse recovery; see [1, 3, 5, 20, 24]. In particular, the range is relevant to non-convex recovery by Schatten quasi-norm minimization. The connection with Gelfand numbers is more than a formal analogy: Chávez-Domínguez and Kutzarova [3] characterized the stability of such recovery procedures in terms of Gelfand widths of suitable Schatten class embeddings. Sharp estimates for these quantities thus describe both the geometry of matrix subspaces and limitations on stable recovery from linear measurements.
The study of general approximation quantities for finite-dimensional sequence spaces has a long history. For Gelfand numbers, fundamental contributions are due to Gluskin [7] and Garnaev and Gluskin [6], whereas the entropy numbers of embeddings between finite-dimensional -spaces were studied by Schütt [30] and subsequently in [4, 32, 10, 19]. In the non-commutative setting, related approximation quantities for embeddings between Schatten classes had already been investigated by Carl and Defant [2] and Hinrichs and Michels [12], in addition to the work of Chávez-Domínguez and Kutzarova [3] discussed above. Hinrichs, Prochno and Vybíral [13] subsequently determined the asymptotic behaviour of the entropy numbers of for all , up to constants depending only on and . However, the corresponding description of Gelfand numbers proved less complete. In [14], the authors obtained asymptotically sharp estimates in a wide range of parameters, extending classical sequence-space results and complementing the earlier bounds. Prochno and Strzelecki [22] subsequently established matching bounds for Gelfand numbers in the intermediate codimension range for and obtained new estimates for approximation and Kolmogorov numbers of Schatten class embeddings. Still, two gaps remained in the description of Gelfand numbers in [22]. One concerns embeddings from , with , into , with : sharp upper bounds were available, but a matching lower estimate throughout the range of indices bounded away from by a fixed proportion had not been established. The purpose of the present paper is to prove the lower estimate proposed in [22 Conjecture 1.4]. Together with the previously known bounds, including the boundary cases and , this completes the asymptotic description of the Gelfand numbers for all indices in the range and .
Presentation of the main results
For and , let denote the space of all real matrices equipped with the Schatten -norm (or quasi-norm when ), defined by
where are the singular values of , that is, the eigenvalues of , counted with multiplicity. In particular, the Schatten -norm coincides with the operator norm of . We write for the natural identity mapping between these spaces. Given a bounded linear operator between quasi-Banach spaces, its -th Gelfand number is defined by
Equivalently, the Gelfand numbers are given by (here and below, all infima over are taken over linear subspaces, which are automatically closed in the finite-dimensional setting)
Throughout, and constants implicit in and are independent of and . Dependence on fixed parameters is indicated by subscripts.
As already mentioned, the purpose of this paper is to determine the Gelfand numbers of in the range and . Hinrichs, Prochno and Vybíral [14 Proposition 4.5] obtained upper estimates for all indices in this range. Prochno and Strzelecki subsequently proved matching lower estimates in the intermediate range of codimensions [22 Proposition 3.1]. For small and moderately large codimensions, the missing estimate was formulated in [22 Conjecture 1.4], which asserts that
here , , and is the fixed universal constant denoted by in [22]. The endpoint was already covered by [22 Proposition 3.2], and the case was also known; see [22 Proposition 3.3] and the references therein. Thus, the missing part of (2) concerns and .
Our main result establishes the conjectured lower estimate for any fixed proportion of the ambient dimension. We include the known matching upper estimate to emphasise that both bounds can be chosen uniformly in the target space index .
Theorem A. Let and . There exist constants and such that, for every , every , and every integer ,
The lower estimate in Theorem A is the contribution of this paper. Taking proves [22 Conjecture 1.4]. The upper estimate is known and follows from the corresponding bound with target space (see [22 Proposition 3.3], with ). In particular, its constant depends only on , whereas the lower-bound constant depends only on and . An explicit choice of is given in (10) below.
Remark 1.1 (The complete range of indices). Combining Theorem A with [14 Proposition 4.5] and [22 Proposition 3.1] yields a complete description of the Gelfand numbers for and . More precisely, with the universal constant introduced above, we have
The estimates in the second range were already sharp. The contribution of Theorem A is the missing matching lower bound in the first range for and . Moreover, in this first range, both comparison constants can be chosen independently of .
Idea of Proof
The proof combines an elementary projection argument with a reduction to rectangular matrices. Let be a linear subspace of dimension . We consider the orthogonal projections onto , with respect to the Hilbert–Schmidt inner product, of the standard basis matrices, each having a single entry equal to one and all other entries equal to zero. Orthogonality relates the corresponding entry of each projected matrix to its squared Hilbert–Schmidt norm. An averaging argument, combined with the rank bound for matrices, then yields
In particular, when occupies a fixed proportion of the rectangular matrix space, this lower bound has order .
To apply this observation, we intersect an arbitrary subspace of codimension less than with the space of matrices supported on the first columns. For , choosing comparable to ensures that the intersection occupies a proportion of this rectangular space bounded below in terms of . The preceding estimate consequently gives the desired factor . The initial range follows from a rank-one argument, while the passage to all uses the trivial estimate . The argument uses neither convexity of the Schatten unit ball nor quasi-Banach duality and therefore applies without modification when .
Organization of the paper
The paper is organized as follows. Section 2 contains the notation and elementary facts about Schatten norms and Gelfand numbers. In Section 3, we establish the rectangular projection estimate and prove Theorem A.
Notation and preliminaries
We present some notation and preliminary facts that are used in the paper.
Matrix spaces and orthogonal projections. For positive integers , write and , and for the trace of a linear operator on a finite-dimensional vector space. We equip with the Hilbert–Schmidt inner product
Let and denote the standard basis vectors of and , respectively, regarded as column vectors. The matrices
each having a single entry equal to one in position and zeros elsewhere, form an orthonormal basis of . For a linear subspace , we write for the orthogonal projection onto with respect to this inner product. Viewed as a linear operator on , it has trace .
Schatten quasi-norms and rank estimates. For , let denote its singular values. We extend the notation from square to rectangular matrices by taking the -norm, or quasi-norm when , of this sequence. In particular, is the operator norm of , and . Throughout, we use the convention .
If , where , then for
The first inequality follows from the monotonicity of sequence quasi-norms. For the second, note that there are at most nonzero singular values. If , Hölder’s inequality with exponents and gives, after padding the nonzero singular-value sequence with zeros,
Taking the -th root proves the assertion. The cases and are immediate. Under the same rank assumption, we shall in particular use that for all
Rectangular matrices as subspaces. For , we identify with the subspace of consisting of matrices whose last columns vanish, via . Since , this identification leaves all nonzero singular values unchanged. Consequently, , . We shall therefore pass between these rectangular and square matrix representations without changing the notation.
Gelfand numbers. For , the condition in (1) is equivalent to . Restricting to a smaller subspace cannot increase the restriction norm. Hence the infimum may be taken over subspaces of dimension exactly , and
For , the full embedding has norm one: the upper bound follows from (4), and equality is attained on any nonzero rank-one matrix. Moreover, applying (4) with gives
This lower bound is attained at the last index. Indeed, let denote the identity matrix and choose in (6) for . All singular values of equal , so for every ,
Together with (7), for this yields
Finally, the inequality implies
Thus, to prove the lower bound in Theorem A, it suffices to consider the target space .
A rectangular projection argument
We first establish the subspace estimate that is the main ingredient of the proof. Its formulation includes the range , although only is needed for Theorem A.
Lemma 3.1. Let , and let be a subspace of dimension . Then, for ,
Proof. Let be the Hilbert–Schmidt-orthogonal projection onto the subspace . Since , we have
As the trace of a finite-dimensional linear operator is independent of the choice of basis and the matrices form an orthonormal basis of , we have
But since is the orthogonal projection onto the -dimensional space , we have and, consequently,
Hence, by averaging, there must exist indices such that
Now put . Since is the orthogonal projection onto , we have
and so . Therefore
Since the absolute value of every matrix entry is bounded by the operator norm,
On the other hand, and hence . By (5),
It follows that
which proves the result.
The parameter determines the dimension of the rectangular matrix space. Intersecting it with a subspace of codimension less than leaves a subspace of dimension at least . We keep free because the resulting lower bound depends on , and a suitable choice of later on will balance the intersection dimension against the rank-dependent norm estimate.
Proposition 3.2. Let , , and . Then
Moreover, whenever .
Proof. Fix a subspace with . For , let denote the subspace of matrices whose last columns vanish. By deleting these zero columns, we identify with . As observed above, this identification preserves all Schatten norms and quasi-norms, and . Now consider the canonical quotient map
Its restriction has kernel . Moreover,
Hence, by the rank–nullity theorem,
Suppose now that . Then . Put . Applying Lemma 3.1 to the subspace , and using , we obtain
Since and for every , it follows that
This estimate holds for every integer satisfying and . We may therefore maximize over all such . Since was arbitrary, taking the infimum over all subspaces with yields
which proves (8).
Finally, suppose that . Taking in (9) gives , and hence . Every nonzero matrix in has rank one, so all of its Schatten norms and quasi-norms coincide. Therefore
Since in the parameter range under consideration, the full embedding has norm one. Consequently, whenever ,
which completes the proof.
We now choose the rectangular dimension in Proposition 3.2 and prove the main result.
Proof of Theorem A. Put . We first prove the lower estimate. The proof uses only , and we distinguish three ranges of indices.
Case 1: . Proposition 3.2 gives . Since , the lower estimate holds with constant one.
Case 2: . Choose
Then , , and
where we used . In particular, . Applying (8), we obtain
Case 3: . Take in (8). Since ,
The inequality implies
Consequently,
Empty ranges are omitted. Combining the three cases proves the lower estimate in (3); one may take
In particular, this constant is independent of .
For the upper estimate, we use the known bound
see [22 Proposition 3.3, with $q=2$]; compare also [14 Proposition 4.5, Step 1]. Since , the inequality holds for every matrix . Taking the restriction norm on each admissible subspace and then the infimum gives
Together with (11), this proves the upper estimate in (3) with a constant independent of both and .
Acknowledgement
The author was supported by the German Research Foundation under DFG project 516672205 Limit theorems for the volume of random projections of -balls. Most of this work was carried out during a research semester at Graz University of Technology. The author gratefully acknowledges the excellent working conditions and the hospitality provided during his stay.
ChatGPT (GPT-5.6 Luna, OpenAI) was used to improve the wording and for grammar and spelling checks, and to assist in the production of Figure 1.

Figure 1. Schematic overview of the Gelfand numbers of ; compare [22 Figure 2(a)]. Light grey indicates parameter regions with previously known sharp asymptotics. Dark grey highlights and , where Theorem A closes the remaining gap and Remark 1.1 gives the resulting all-index asymptotics. The hatched triangle corresponds to , where matching bounds are still missing for intermediate indices . The boundary of the two highlighted regions is not part of the open cases: sharp asymptotic estimates were already known for with , for with , for with , and along the diagonal .
References
- [1]E. J. Candès and B. Recht. Exact matrix completion via convex optimization. Found. Comput. Math., 9(6):717–772, 2009.
- [2]B. Carl and A. Defant. Asymptotic estimates for approximation quantities of tensor product identities. J. Approx. Theory, 88(2):228–256, 1997.DOI
- [3]J. A. Chávez-Domínguez and D. Kutzarova. Stability of low-rank matrix recovery and its connections to Banach space geometry. J. Math. Anal. Appl., 427(1):320–335, 2015.
- [4]D. E. Edmunds and H. Triebel. Function spaces, entropy numbers, differential operators, volume 120 of Cambridge Tracts in Mathematics. Cambridge University Press, Cambridge, 1996.
- [5]S. Foucart and H. Rauhut. A mathematical introduction to compressive sensing. Applied and Numerical Harmonic Analysis. Birkhäuser/Springer, New York, 2013.DOI
- [6]A. Y. Garnaev and E. D. Gluskin. The widths of a Euclidean ball. Soviet Math. Dokl., 30:200–204, 1984.
- [7]E. D. Gluskin. Norms of random matrices and diameters of finite-dimensional sets. Mat. Sb. (N.S.), 120(162) (2):180–189, 286, 1983.DOI
- [8]Y. Gordon and D. R. Lewis. Absolutely summing operators and local unconditional structures. Acta Math., 133:27–48, 1974.DOI
- [9]A. Grothendieck. Sur une notion de produit tensoriel topologique d’espaces vectoriels topologiques, et une classe remarquable d’espaces vectoriels liée à cette notion. C. R. Acad. Sci. Paris, 233:1556–1558, 1951.
- [10]O. Guédon and A. E. Litvak. Euclidean projections of a p-convex body. In Geometric aspects of functional analysis, volume 1745 of Lecture Notes in Math., pages 95–108. Springer, Berlin, 2000.DOI
- [11]O. Guédon and G. Paouris. Concentration of mass on the Schatten classes. Ann. Inst. H. Poincaré Probab. Statist., 43(1):87–99, 2007.DOI
- [12]A. Hinrichs and C. Michels. Approximation numbers of inclusions between Schatten classes. Rend. Circ. Mat. Palermo (2) Suppl., (76):395–411, 2005.
- [13]A. Hinrichs, J. Prochno, and J. Vybíral. Entropy numbers of embeddings of Schatten classes. J. Funct. Anal., 273(10):3241–3261, 2017.arxiv.org/abs/1612.08105
- [14]A. Hinrichs, J. Prochno, and J. Vybíral. Gelfand numbers of embeddings of Schatten classes. Math. Ann., 380(3–4):1563–1593, 2021.arxiv.org/abs/2011.06554
- [15]Z. Kabluchko, J. Prochno, and C. Thäle. Exact asymptotic volume and volume ratio of Schatten unit balls. J. Approx. Theory, 257:105457, 2020.arxiv.org/abs/1804.03467
- [16]Z. Kabluchko, J. Prochno, and C. Thäle. Intersection of unit balls in classical matrix ensembles. Israel J. Math., 239(1):129–172, 2020.arxiv.org/abs/1804.03466
- [17]Z. Kabluchko, J. Prochno, and C. Thäle. Sanov-type large deviations in Schatten classes. Ann. Inst. Henri Poincaré Probab. Stat., 56(2):928–953, 2020.arxiv.org/abs/1808.04862
- [18]H. König, M. Meyer, and A. Pajor. The isotropy constants of the Schatten classes are bounded. Math. Ann., 312(4):773–783, 1998.DOI
- [19]T. Kühn. A lower estimate for entropy numbers. J. Approx. Theory, 110(1):120–124, 2001.DOI
- [20]C. Kümmerle and J. Sigl. Harmonic mean iteratively reweighted least squares for low-rank matrix recovery. J. Mach. Learn. Res., 19:Paper no. 47, 49, 2018.arxiv.org/abs/1703.05038
- [21]S. Kwapień and A. Pełczyński. The main triangle projection in matrix spaces and its applications. Studia Math., 34:43–68, 1970.DOI
- [22]J. Prochno and M. Strzelecki. Approximation, Gelfand, and Kolmogorov numbers of Schatten class embeddings. J. Approx. Theory, 277:105736, 2022.DOI
- [23]J. Radke and B.-H. Vritsiou. On the thin-shell conjecture for the Schatten classes. Ann. Inst. H. Poincaré Probab. Statist., 56(1):87–119, 2020.arxiv.org/abs/1602.06934
- [24]A. Rohde and A. B. Tsybakov. Estimation of high-dimensional low-rank matrices. Ann. Statist., 39(2):887–930, 2011.
- [25]A. F. Ruston. On the Fredholm theory of integral equations for operators belonging to the trace class of a general Banach space. Proc. London Math. Soc. (2), 53:109–124, 1951.DOI
- [26]R. Schatten. The cross-space of linear transformations. Ann. of Math. (2), 47:73–84, 1946.DOI
- [27]R. Schatten. Norm ideals of completely continuous operators. Ergebnisse der Mathematik und ihrer Grenzgebiete. N. F., Heft 27. Springer-Verlag, Berlin-Göttingen-Heidelberg, 1960.
- [28]R. Schatten and J. von Neumann. The cross-space of linear transformations. II. Ann. of Math. (2), 47:608–630, 1946.DOI
- [29]R. Schatten and J. von Neumann. The cross-space of linear transformations. III. Ann. of Math. (2), 49:557–582, 1948.DOI
- [30]C. Schütt. Entropy numbers of diagonal operators between symmetric Banach spaces. J. Approx. Theory, 40(2):121–128, 1984.DOI
- [31]N. Tomczak-Jaegermann. The moduli of smoothness and convexity and the Rademacher averages of trace classes S_p (1 ≤ p < ∞). Studia Math., 50:163–182, 1974.
- [32]H. Triebel. Fractals and spectra, volume 91 of Monographs in Mathematics. Birkhäuser Verlag, Basel, 1997. Related to Fourier analysis and function spaces.