Showing 1–9 of 9 results for author: Gutin, R

A conjecture concerning *-algebras that unifies some matrix decompositions

Authors: Ran Gutin

Abstract: In this note, we propose a simple-looking but broad conjecture about star-algebras over the field of real numbers. The conjecture enables many matrix decompositions to be represented by star-algebras and star-ideals. This paper is written for people with a background in representation theory and module theory. The motivation for investigating this is the possibility of expressing polymorphic algor… ▽ More In this note, we propose a simple-looking but broad conjecture about star-algebras over the field of real numbers. The conjecture enables many matrix decompositions to be represented by star-algebras and star-ideals. This paper is written for people with a background in representation theory and module theory. The motivation for investigating this is the possibility of expressing polymorphic algorithms in numerical and theoretical linear algebra. This is similar to but different from algebraic (semiring based) approaches to dynamic programming. We prove certain cases of the conjecture. △ Less

Submitted 8 August, 2023; originally announced August 2023.

An interactive visualisation for all 2x2 real matrices, with applications to conveying the dynamics of iterative eigenvalue algorithms

Authors: Ran Gutin

Abstract: We present two interactive visualisations of 2x2 real matrices, which we call v1 and v2. v1 is only valid for PSD matrices, and uses the spectral theorem in a trivial way -- we use it as a warm-up. By contrast, v2 is valid for *all* 2x2 real matrices, and is based on the lesser known theory of Lie Sphere Geometry. We show that the dynamics of iterative eigenvalue algorithms can be illustrated usin… ▽ More We present two interactive visualisations of 2x2 real matrices, which we call v1 and v2. v1 is only valid for PSD matrices, and uses the spectral theorem in a trivial way -- we use it as a warm-up. By contrast, v2 is valid for *all* 2x2 real matrices, and is based on the lesser known theory of Lie Sphere Geometry. We show that the dynamics of iterative eigenvalue algorithms can be illustrated using both. v2 has the advantage that it simultaneously depicts many properties of a matrix, all of which are relevant to the study of eigenvalue algorithms. Examples of the properties of a matrix that v2 can depict are its Jordan Normal Form and orthogonal similarity class, as well as whether it is triangular, symmetric or orthogonal. Despite its richness, using v2 interactively seems rather intuitive. △ Less

Submitted 3 November, 2022; originally announced November 2022.

Unitary canonical forms over Clifford algebras, and an observed unification of some real-matrix decompositions

Authors: Ran Gutin

Abstract: We show that the spectral theorem -- which we understand to be a statement that every self-adjoint matrix admits a certain type of canonical form under unitary similarity -- admits analogues over other $*$-algebras distinct from the complex numbers. If these $*$-algebras contain nilpotents, then it is shown that there is a consistent way in which many classic matrix decompositions -- such as the S… ▽ More We show that the spectral theorem -- which we understand to be a statement that every self-adjoint matrix admits a certain type of canonical form under unitary similarity -- admits analogues over other $*$-algebras distinct from the complex numbers. If these $*$-algebras contain nilpotents, then it is shown that there is a consistent way in which many classic matrix decompositions -- such as the Singular Value Decomposition, the Takagi decomposition, the skew-Takagi decomposition, and the Jordan decomposition, among others -- are immediate consequences of these. If producing the relevant canonical form of a self-adjoint matrix were a subroutine in some programming language, then the corresponding classic matrix decomposition would be a 1-line invocation with no additional steps. We also suggest that by employing operator overloading in a programming language, a numerical algorithm for computing a unitary diagonalisation of a complex self-adjoint matrix would generalise immediately to solving problems like SVD or Takagi. While algebras without nilpotents (like the quaternions) allow for similar unifying behaviour, the classic matrix decompositions which they unify are never obtained as easily. In the process of doing this, we develop some spectral theory over Clifford algebras of the form $\cl_{p,q,0}(\mathbb R)$ and $\cl_{p,q,1}(\mathbb R)$ where the former is admittedly quite easy. We propose a broad conjecture about spectral theorems. △ Less

Submitted 24 January, 2023; v1 submitted 8 August, 2022; originally announced August 2022.

A visualisation for conveying the dynamics of iterative eigenvalue algorithms over PSD matrices

Authors: Ran Gutin

Abstract: We propose a new way of visualising the dynamics of iterative eigenvalue algorithms such as the QR algorithm, over the important special case of PSD (positive semi-definite) matrices. Many subtle and important properties of such algorithms are easily found this way. We believe that this may have pedagogical value to both students and researchers of numerical linear algebra. The fixed points of ite… ▽ More We propose a new way of visualising the dynamics of iterative eigenvalue algorithms such as the QR algorithm, over the important special case of PSD (positive semi-definite) matrices. Many subtle and important properties of such algorithms are easily found this way. We believe that this may have pedagogical value to both students and researchers of numerical linear algebra. The fixed points of iterative algorithms are obtained visually, and their stability is analysed intuitively. It becomes clear that what it means for an iterative eigenvalue algorithm to "converge quickly" is an ambiguous question, depending on whether eigenvalues or eigenvectors are being sought. The presentation is likely a novel one, and using it, a theorem about the dynamics of general iterative eigenvalue algorithms is proved. There is an accompanying video series, currently hosted on Youtube, that has certain advantages in terms of fully exploiting the interactivity of the visualisation. △ Less

Submitted 31 March, 2022; originally announced April 2022.

Towards a Singular Value Decomposition and spectral theory for all rings

Authors: Ran Gutin

Abstract: We propose definitions of SVD, spectral decomposition (for self-adjoint matrices) and Jordan decomposition which make sense for all rings. For many rings, these decompositions can be shown to exist. For some specific rings, these decompositions are complicated to describe in full and prove the existence of. These decompositions have occurred piecemeal in the literature. We conjecture that they exi… ▽ More We propose definitions of SVD, spectral decomposition (for self-adjoint matrices) and Jordan decomposition which make sense for all rings. For many rings, these decompositions can be shown to exist. For some specific rings, these decompositions are complicated to describe in full and prove the existence of. These decompositions have occurred piecemeal in the literature. We conjecture that they exist for many rings, including all Clifford algebras over the real numbers and complex numbers. The origin of this programme is not directly in module theory or linear algebra. △ Less

Submitted 19 December, 2021; originally announced December 2021.

An analogue of the relationship between SVD and pseudoinverse over double-complex matrices

Authors: Ran Gutin

Abstract: We present a generalisation of the pseudoinverse operation to pairs of matrices, as opposed to single matrices alone. We note the fact that the Singular Value Decomposition can be used to compute the ordinary Moore-Penrose pseudoinverse. We present an analogue of the Singular Value Decomposition for pairs of matrices, which we show is inadequate for our purposes. We then present a more sophisticat… ▽ More We present a generalisation of the pseudoinverse operation to pairs of matrices, as opposed to single matrices alone. We note the fact that the Singular Value Decomposition can be used to compute the ordinary Moore-Penrose pseudoinverse. We present an analogue of the Singular Value Decomposition for pairs of matrices, which we show is inadequate for our purposes. We then present a more sophisticated analogue of the SVD which includes features of the Jordan Normal Form, which we show is adequate for our purposes. This analogue of the SVD, which we call the Jordan SVD, was already presented in a previous paper by us called "Matrix decompositions over the double numbers". We adopt the idea presented in the same paper that a pair of matrices is actually a single matrix over the double number system. △ Less

Submitted 6 December, 2021; v1 submitted 21 May, 2021; originally announced May 2021.

Comments: arXiv admin note: text overlap with arXiv:2105.08047

Matrix decompositions over the double numbers

Authors: Ran Gutin

Abstract: We take matrix decompositions that are usually applied to matrices over the real numbers or complex numbers, and extend them to matrices over an algebra called the double numbers. In doing so, we unify some matrix decompositions: For instance, we reduce the LU decomposition of real matrices to LDL decomposition of double matrices; we similarly reduce eigendecomposition of real matrices to singular… ▽ More We take matrix decompositions that are usually applied to matrices over the real numbers or complex numbers, and extend them to matrices over an algebra called the double numbers. In doing so, we unify some matrix decompositions: For instance, we reduce the LU decomposition of real matrices to LDL decomposition of double matrices; we similarly reduce eigendecomposition of real matrices to singular value decomposition of double matrices. Notably, these are opposite to the usual reductions. This provides insight into linear algebra over the familiar real numbers and complex numbers. We also show that algorithms that are valid for complex matrices are often equally valid for double matrices. We finish by proposing a new matrix decomposition called the Jordan SVD, which we use to challenge a claim made in Yaglom's book Complex Numbers In Geometry concerning Linear Fractional Transformations over the double numbers. △ Less

Submitted 6 December, 2021; v1 submitted 17 May, 2021; originally announced May 2021.

Comments: arXiv admin note: text overlap with arXiv:2105.10455

Generalisations of Singular Value Decomposition to dual-numbered matrices

Authors: Ran Gutin

Abstract: We present two generalisations of Singular Value Decomposition from real-numbered matrices to dual-numbered matrices. We prove that every dual-numbered matrix has both types of SVD. Both of our generalisations are motivated by applications, either to geometry or to mechanics. We present two generalisations of Singular Value Decomposition from real-numbered matrices to dual-numbered matrices. We prove that every dual-numbered matrix has both types of SVD. Both of our generalisations are motivated by applications, either to geometry or to mechanics. △ Less

Submitted 9 June, 2021; v1 submitted 19 July, 2020; originally announced July 2020.

Constructive proof of Herschfeld's Convergence Theorem

Authors: Ran Gutin

Abstract: In this paper, we present a constructive proof of Herschfeld's Convergence Theorem. Our formulation differs from Herschfeld's in a few ways: We consider radicals that nest transfinitely many times, as these are essential to the proof; additionally, we formulate the conditions for convergence in such a way that a constructive proof is possible. In this paper, we present a constructive proof of Herschfeld's Convergence Theorem. Our formulation differs from Herschfeld's in a few ways: We consider radicals that nest transfinitely many times, as these are essential to the proof; additionally, we formulate the conditions for convergence in such a way that a constructive proof is possible. △ Less

Submitted 22 May, 2020; v1 submitted 5 July, 2019; originally announced July 2019.