Classification of abelian finite-dimensional C โ C^{*} -algebras by orthogonality
Bojan Kuzma, Sushil Singla ยท 2024
Read Classification of abelian finite-dimensional C โ C^{*} -algebras by orthogonality by Bojan Kuzma, Sushil Singla (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2411-01684.txt.
Description
A 2024 arXiv research paper by Bojan Kuzma, Sushil Singla. The main goal of the article is to prove that if ๐ 1 \mathcal{A}_{1} and ๐ 2 \mathcal{A}_{2} are Birkhoff-James isomorphic C โ C^{*} -algebras over the fields ๐ฝ 1 \mathbb{F}_{1} and ๐ฝ 2 \mathbb{F}_{2} , respectively and if ๐ 1 \mathcal{A}_{1} finite-dimensional, abelian of dimension greater than one, then ๐ฝ 1 = ๐ฝ 2 \mathbb{F}_{1}=\mathbb{F}_{2} and ๐ 1 \mathcal{A}_{1} and ๐ 2 \mathcal{A}_{2} are (isometrically) โ \ast -isomorphic C โ C^{*} -algebras.
Catalog record
| CREATOR | Bojan Kuzma, Sushil Singla |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 71 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2411-01684
Browse the full library ยท About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Classification of abelian finite-dimensional C โ C^{*} -algebras by orthogonality Bojan Kuzma, Sushil Singla Abstract . The main goal of the article is to prove that if ๐ 1 \mathcal{A}_{1} and ๐ 2 \mathcal{A}_{2} are Birkhoff-James isomorphic C โ C^{*} -algebras over the fields ๐ฝ 1 \mathbb{F}_{1} and ๐ฝ 2 \mathbb{F}_{2} , respectively and if ๐ 1 \mathcal{A}_{1} finite-dimensional, abelian of dimension greater than one, then ๐ฝ 1 = ๐ฝ 2 \mathbb{F}_{1}=\mathbb{F}_{2} and ๐ 1 \mathcal{A}_{1} and ๐ 2 \mathcal{A}_{2} are (isometrically) โ \ast -isomorphic C โ C^{*} -algebras. Furthermore, it is also proved that for a finite-dimensional C โ C^{*} -algebra ๐ {\mathcal{A}} , we have โ ๐ โฅ \mathcal{L}_{{\mathcal{A}}}^{\bot} is the sum of minimal ideals which are not skew-fields and โ ๐ โฅ โฃ โฅ \mathcal{L}_{{\mathcal{A}}}^{\bot\bot} is the sum of minimal ideals which are skew-fields, where โ ๐ \mathcal{L}_{\mathcal{A}} denotes the set of all left-symmetric elements in ๐ \mathcal{A} and for any subset ๐ฎ โ ๐ \mathcal{S}\subseteq\mathcal{A} , the set ๐ฎ โฅ \mathcal{S}^{\bot} represents the set of all elements of ๐ \mathcal{A} which are Birkhoff-James orthogonal to ๐ฎ \mathcal{S} . A procedure to extract the minimal ideals which are (commutative) fields is also given. 1991 Mathematics Subject Classification Primary 46L05; Secondary 46B20, 46B80 1. Introduction By Gelfand transform every unital abelian complex C^{*} -algebra \mathcal{A} is \ast -isomorphic to C(X) , the space of complex-valued continuous functions on some compact Hausdorff space. This translates the study of algebraic properties to the study of topological properties (and vice-versa): C(X) is \ast -isomorphic to C(Y) if and only if X and Y are homeomorphic topological spaces. Formally, Gelfand transform is a contravariant equivalence between the category of unital abelian C^{*} -algebras and the category of the space of compact Hausdorff spaces (see, e.g., [ 2 , 5 , 6 ] for more information). Recently Tanaka [ 23 , Theorem 3.5, Corollary 3.6] showed that the same can be achieved by studying the geometrical properties rather than the topological ones: He characterized abelian complex C^{*} -algebras among all complex C^{*} -algebras by using only the underlying geometric structure. Moreover, he showed that two complex abelian C^{*} -algebras are \ast -isomorphic if and only if their geometric structures are homeomorphic. The geometric structure was initially defined in terms of Birkhoff-James orthogonality (see [ 22 , Definition 3.4] ). Figuratively speaking, suppose we obtain a cast of the closed unit ball of the C^{\ast} -algebra norm. We are allowed to examine it with a sufficiently long, infinitesimally thin needle by placing it tangentially in various directions onto the unit sphere of the norm, then translating it parallelly to the center of the ball and examining the points which the translated needle cuts out from the boundary. If x is the touching point of the needle and y is the cut-out point of the translated needle, then the tangentiality of the needle at point x in direction y is equivalent to \|x+\lambda y\|\geq\|x\| for each scalar \lambda , that is, to Birkhoff-James orthogonality of x and y (for complex C^{\ast} -algebras, the probing needle has two real dimensions). The geometric structure was defined in terms of maximal faces of C^{*} -normโs unit ball and requires no knowledge of algebraic operations. Following the approach outlined above we, in [ 17 , Theorem 2.1] , completely classified the objects in the categories of real or complex finite-dimensional simple C^{*} -algebras by using only the relation of Birkhoff-James orthogonality, that is, by relying only on the norm structure alone. Another important aspect of our work was that we worked with real as well as complex C^{*} -algebras simultaneously and even gave a procedure to characterize the underlying field when the dimension of space is greater than one. The theory of real C^{*} -algebras is similar to complex C^{*} -algebras, see [ 4 , 6 , 18 , 21 ] (and [ 19 ] for a review of their applications) though being able to characterize the underlying field of a given real or complex C^{*} -algebra, from Birkhoff-James orthogonality alone, was still a bit surprising. In this article we continue our work in the categories of finite-dimensional real or complex C^{*} -algebras and characterize (pseudo-)abelian C^{*} -algebras together with the underlying fields when the dimension of C^{*} -algebra is greater than one. A few notations are in order. In the sequel, \mathcal{A} will stand for a finite-dimensional C^{*} -algebra over the field \mathbb{F}\in\{\mathbb{R},\mathbb{C}\} . We will denote the matrix block decomposition of a complex C^{*} -algebra \mathcal{A} by (1.1) \mathcal{M}_{n_{1}}(\mathbb{C})\oplus\dots\oplus\mathcal{M}_{n_{\ell}}(\mathbb{C}) for some positive integers n_{1}\leq\dots\leq n_{\ell} . Similarly, the matrix block decomposition of a real C^{*} -algebra \mathcal{A} will be denoted by (1.2) \mathcal{M}_{n_{1}}(\mathbb{K}_{1})\oplus\dots\oplus\mathcal{M}_{n_{\ell}}(\mathbb{K}_{\ell}) where \mathbb{K}_{i}\in\{\mathbb{R},\mathbb{C},\mathbb{H}\} with real dimension d_{1},\dots,d_{\ell} such that d_{i}\leq\dots\leq d_{j} whenever n_{i}=\dots=n_{j} (every {\mathcal{A}} has such decomposition, see [ 6 , Theorem 1.5, Theorem 8.4] for more information). Recall that a non-zero two-sided ideal of {\mathcal{A}} is minimal if it does not properly contain any other non-trivial two-sided ideal. We will refer to the sum of those minimal ideals of {\mathcal{A}} which are skew-fields as a pseudo-abelian summand of {\mathcal{A}} and we will refer to the sum of abelian minimal ideals as an abelian summand of {\mathcal{A}} . If C^{*} -algebra is decomposed as in ( 1.1 ) or ( 1.2 ), then its minimal ideals coincide with the individual blocks, its pseudo-abelian summand coincides with the sum of all blocks of size one, and its abelian summand coincides with the sum of all blocks of size one over the real or complex field; see, e.g., Wedderburn-Artin theorem [ 7 , V.4.6] . Therefore, if \mathbb{F}=\mathbb{C} the pseudo-abelian summand coincides with the abelian summand of {\mathcal{A}} . Also, if {\mathcal{A}} is an abelian C^{*} -algebra, then its pseudo-abelian summand equals {\mathcal{A}} . However, in case {\mathbb{F}}={\mathbb{R}} the pseudo-abelian summand might contain a quaternionic block in which case it is not abelian. We will say that \mathcal{A} is a pseudo-abelian C^{*} -algebra if it equals to its pseudo-abelian summand. In the case when {\mathbb{F}}={\mathbb{C}} this is the same as an abelian C^{*} -algebra. Let us briefly discuss also the definition and basic properties of Birkhoff-James (BJ) orthogonality. For two vectors v,w in a normed space V over a field \mathbb{F} , v is said to be BJ orthogonal to w , denoted by v\perp w if \|v\|\leq\|v+\lambda w\|\text{ for all }\lambda\in{\mathbb{F}}. One can easily see that this relation is homogeneous and that, equivalently, v\perp w if and only if f_{v}(w)=0 for some supporting functional f_{v} at v (that is, \|f_{v}\|=1 and f_{v}(v)=\|v\| ), see [ 13 , Theorem 2.1] or [ 8 , Equation 2.1] . Therefore, if we define the outgoing neighborhood of v by (1.3) v^{\bot}=\{w\in V;\;\;v\perp w\}, then we have (1.4) v^{\bot}=\bigcup\{\mathop{\mathrm{ker}}f;\;\;f\text{ is supporting functional at }v\}. A bijective map \phi\colon V\rightarrow V^{\prime} is a BJ isomorphism between V and V^{\prime} if v\perp w\iff\phi(v)\perp\phi(w)\quad\text{ for all }v,w\in V. Two normed spaces V and V^{\prime} are BJ isomorphic if there exists a BJ isomorphism between them. We can now state our main results. Recall that a finite-dimensional C^{*} -algebra is pseudo-abelian if it contains contains only blocks of size one in its matrix block decomposition. Theorem 1.1 . Let {\mathcal{A}}_{1} and {\mathcal{A}}_{2} be