DOCSLIB.ORG

Hypercomplex Principal Component Pursuit Via Convex Optimization

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Hypercomplex Principal Component Pursuit Via Convex Optimization Proceedings, APSIPA Annual Summit and Conference 2018 12-15 November 2018, Hawaii Hypercomplex Principal Component Pursuit via Convex Optimization Takehiko Mizoguchi∗y and Isao Yamada∗ ∗Department of Information and Communications Engineering, Tokyo Institute of Technology, Tokyo, Japan yNEC Corporation, Kanagawa, Japan E-mail: fmizoguchi,[email protected] Abstract—Expressing multidimensional information as a value color, material, intensity, etc., and each attribute may have in hypercomplex number systems (e.g., quaternion, octonion, etc.) correlation with other attributes. Modeling of the correlations has great potential, in signal processing applications, to enjoy among attributes in multidimensional data and will be more their nontrivial algebraic benefits which are not available in standard real or complex vector systems. Exploiting sparsity of and more important by the popularization of 3D printer [11], hypercomplex matrices or vectors would play an important role virtual reality, medical imaging etc. Algebraically natural oper- in utilizing such benefits. In this paper, we propose a new spar- ations in hypercomplex number system has great potential for sity measure for evaluating sparsity of hypercomplex matrices. such modelingof various correlations (see e.g., [12]–[16] for With utilizing this measure, the hypercomplex robust principal color image processing applications). However, because of the component analysis can be relaxed into the hypercomplex robust principal component pursuit and it can be reduced to a real “singularity” of higher dimensional C-D number systems (see convex optimization problem. We then derive an algorithmic e.g., Example 1), few mathematical tools had been maintained solution to the hypercomplex principal component pursuit based [17]–[20]. To overcome this situation, in our previous works on a proximal splitting technique. Numerical experiments are [21]–[23], we have proposed several useful mathematical tools performed in the octonion domain and show that the proposed for designing advanced algorithm for optimization, learning algorithm outperforms a part-wise state-of-art real, complex and quaternion algorithm. and low rank approximation in hypercomplex domain. In [21] we proposed an algebraic real translation for clarifying the I. INTRODUCTION relation between C-D linear system and real vector valued Multidimensional information arises naturally in many areas linear systems, and successfully designed some online learning of engineering and science since almost all observations have algorithms which are available in general C-D domain. More- many attributes. Utilizing hypercomplex number system for over, we also proposed in [22], [23] useful tools C-D singular representing such multidimensional information is one of the value decomposition, R-rank (see Section II-C) and low rank most effective ways because we can express multidimensional approximation technique and proposed an algorithmic solution information not in terms of vectors but in terms of num- for low rank hypercomplex tensor completion. bers among which we can define the four basic arithmetic One of the other standard approximation methods utilizing operators. Indeed, it has been used in many areas such as simpler structure of matrices is the robust principal component computer graphics [1] and robotics [2], [3] wind forecasting analysis (RPCA) [24], which separates an input matrix into a [4]–[6] and noise reduction in acoustic systems [7]. In the low-rank and sparse ones: statistical signal processing field, effective utilization of the k k m-dimensional Cayley-Dickson number system (C-D number minimize rank(L) + λ S 0 s.t. M = L + S; (1) L;S2RM×N system) [8], [9], which is a standard class of hypercomplex ∥·∥ number systems [10], including, e.g., real R, complex C, where 0 denotes the `0-norm, which counts the number quaternion H, octonion O and sedenion S etc., have been of non-zero entries of S. The RPCA has been successfully investigated. used for various signal processing applications such as source A hypercomplex number has one real part and many imag- separation [25], face recognition [26] and so on. Unfortunately, inary parts, and it can represent multidimensional data as a the formulation (1) is NP-hard, so its convex relaxation called number for which the four arithmetic operations including the principal component pursuit [24] is often considered in multiplication and division are available. It can fulfill the stead of the RPCA. Recently, the PCP is extended to simpler four arithmetic operation for multidimensional information, hypercomplex domains C and H using the complex matrix which are not available for ordinary real multidimensional isomorphism [27]. However, it can be only applied up to four vectors. Moreover, thanks to the nontrivial algebraic struc- dimensional data and hard to be generalized to octonion and ture, the multiplication of hypercomplex numbers can enjoy higher dimensional C-D domain only with this isomorphism. algebraically interactions among real and imaginary parts. In this paper, to establish a RPCA framework in hyper- Hypercomplex vectors, matrices and tensors can also enjoy complex domain, first, we present Cayley-Dickson principal these benefits. For example, in 3D object modeling, each point component analysis (C-D PCP) as a convex relaxation of the in 3-dimensional space can have multiple attribute such as C-D extension of RPCA. This relaxation is based on the 978-988-14768-5-2 ©2018 APSIPA 579 APSIPA-ASC 2018 Proceedings, APSIPA Annual Summit and Conference 2018 12-15 November 2018, Hawaii R-rank proposed in [22] and a new sparsity measure, `1- m-dimensional hypercomplex number Am is extended to A2m norm of C-D matrices. This sparsity measure can be inter- [8], [9] as preted as evaluating a group sparsity of real matrices, so the z := x + yi 2 A ; x; y 2 A ; proximity operator can be easily calculated. Hence, the C-D m+1 2m m PCP is a convex optimization in real domain which can be where im+1 62 Am is the additional imaginary unit for dou- A 2 − solved by applying proximal splitting techniques to a certain bling the dimension of m satisfying im+1 = 1, i1im+1 = structured convex optimization problem. We finally propose im+1i1 = im+1 and ivim+1 = −im+1iv =: im+v for all A an hypercomplex PCP algorithm m based on Douglas- v = 2; : : : ; m. For example, the real number system (A1 :=) R Rachford splitting (DRS) [28]. The proposed algorithm is a is extended into complex number system C (= A2) by the C-D C-D generalization of a PCP algorithm (DR-PCP) proposed construction. Note that the value of m is restricted to the form in [29] and can be applied to general C-D domains. of 2n (n 2 N). The hypercomplex number systems constructed Numerical experiments are performed in the context of inductively from the real number by the C-D construction are recovering sparsely corrupted low rank matrices in octonion called Cayley-Dickson number system (C-D number system). domain and show that the proposed algorithm successfully uti- The imaginary units appeared in the C-D number systems 2 − lizes algebraically natural correlations of real and all imaginary have many characteristic properties [21] such as iα = 1 − 6 2 f g parts to recover much more faithfully the original matrices, and iαiβ = iβiα(α = β)Pfor all α; β 2; : : : ; m . These corrupted randomly by noise, than a part-wise real, complex ∗ m 2 ≥ 2 A properties ensures aa = `=1 a` 0 for any a m in ∗ and quaternion PCP algorithms. (2) and a 2 Am in (3) and enablep us to define the absolute j j ∗ II. PRELIMINARIES values of C-D number a as a := aa . A. Hypercomplex Number System Example 1. 1) A representative example of hypercomplex Let N and R be respectively the set of all non-negative inte- number is the quaternion H. The quaternion number gers and the set of all real numbers. Define an m-dimensional system is constructed from the complex number system hypercomplex number Am (m 2 N n f0g) expanded on the by using the C-D construction. A quaternion number is real vector space [8] a 4-dimensional hypercomplex which is defined as 2 H 2 R a := a1i1 + a2i2 + ··· + amim 2 Am; a1; : : : ; am 2 R (2) q = q1 + q2{ + q3| + q4κ ; q1; q2; q3; q4 based on imaginary units i1;:::; im, where i1 = 1 represents with the multiplication table: the vector identity element. Any hypercomplex number is {| = −|{ = κ, ȷκ = −κȷ = {; κı = −ıκ = |; (4) expressed uniquely in the form of (2). The coefficient of each {2 = |2 = κ2 = −1 imaginary unit a` (` = 1; : : : ; m) is represented as a` = =`(a). A multiplication table defines the products of any imaginary by letting m = 4, i1 = 1, i2 = {, i3 = | and i4 = κ. 2 2 − From (4), quaternions are not commutative, i.e., pq =6 qp unit with each other or with itself (e.g., i1 = 1; i2 = 1 for p; q 2 H in general. and i1i2 = i2i1 = i2 for A2(=: C)). The addition and the subtraction of two hypercomplex numbers are defined as 2) The octonion O can be constructed from the quaternion commutative binomial operations H by the C-D construction. Note that the multiplication in O is neither commutative nor associative, i.e., pq =6 qp ··· a b := (a1 b1)i1 + (a2 b2)i2 + + (am bm)im and (pq)r =6 p(qr) for p; q; r 2 O in general [10]. For the octonion multiplication table, see, e.g., [10]. for a; b 2 Am, where b := b1i1 + b2i2 + ··· + bmim, 2 R b1; : : : ; bm . From the unique expression of (2), the The C-D number system can be seen as an algebraically nat- multiplication of two hypercomplex numbers ural higher dimensional generalization of our familiar fields, R C ab = (a1i1 + a2i2 + ··· + amim)(b1i1 + b2i2 + ··· + bmim) i.e., and . AN f >j 2 A Xm Xm We also define m := [x1; : : : ; xN ] xi m (i = > := akb`iki` 2 Am 1;:::;N)g for 8N 2 N n f0g, where (·) stands for the H N k=1 `=1 transpose.
Load more

Recommended publications
  • Characterizations of Several Split Regular Functions on Split Quaternion in Clifford Analysis
    East Asian Math. J. Vol. 33 (2017), No. 3, pp. 309{315 http://dx.doi.org/10.7858/eamj.2017.023 CHARACTERIZATIONS OF SEVERAL SPLIT REGULAR FUNCTIONS ON SPLIT QUATERNION IN CLIFFORD ANALYSIS Han Ul Kang, Jeong Young Cho, and Kwang Ho Shon* Abstract. In this paper, we investigate the regularities of the hyper- complex valued functions of the split quaternion variables. We define several differential operators for the split qunaternionic function. We re- search several left split regular functions for each differential operators. We also investigate split harmonic functions. And we find the correspond- ing Cauchy-Riemann system and the corresponding Cauchy theorem for each regular functions on the split quaternion field. 1. Introduction The non-commutative four dimensional real field of the hypercomplex num- bers with some properties is called a split quaternion (skew) field S. Naser [12] described the notation and the properties of regular functions by using the differential operator D in the hypercomplex number system. And Naser [12] investigated conjugate harmonic functions of quaternion variables. In 2011, Koriyama et al. [9] researched properties of regular functions in quater- nion field. In 2013, Jung et al. [1] have studied the hyperholomorphic functions of dual quaternion variables. And Jung and Shon [2] have shown hyperholo- morphy of hypercomplex functions on dual ternary number system. Kim et al. [8] have investigated regularities of ternary number valued functions in Clif- ford analysis. Kang and Shon [3] have developed several differential operators for quaternionic functions. And Kang et al. [4] researched some properties of quaternionic regular functions. Kim and Shon [5, 6] obtained properties of hyperholomorphic functions and hypermeromorphic functions in each hyper- compelx number system.
    [Show full text]
  • Notices of the American Mathematical Society
    ISSN 0002-9920 of the American Mathematical Society February 2006 Volume 53, Number 2 Math Circles and Olympiads MSRI Asks: Is the U.S. Coming of Age? page 200 A System of Axioms of Set Theory for the Rationalists page 206 Durham Meeting page 299 San Francisco Meeting page 302 ICM Madrid 2006 (see page 213) > To mak• an antmat•d tub• plot Animated Tube Plot 1 Type an expression in one or :;)~~~G~~~t;~~i~~~~~~~~~~~~~:rtwo ' 2 Wrth the insertion point in the 3 Open the Plot Properties dialog the same variables Tl'le next animation shows • knot Plot 30 Animated + Tube Scientific Word ... version 5.5 Scientific Word"' offers the same features as Scientific WorkPlace, without the computer algebra system. Editors INTERNATIONAL Morris Weisfeld Editor-in-Chief Enrico Arbarello MATHEMATICS Joseph Bernstein Enrico Bombieri Richard E. Borcherds Alexei Borodin RESEARCH PAPERS Jean Bourgain Marc Burger James W. Cogdell http://www.hindawi.com/journals/imrp/ Tobias Colding Corrado De Concini IMRP provides very fast publication of lengthy research articles of high current interest in Percy Deift all areas of mathematics. All articles are fully refereed and are judged by their contribution Robbert Dijkgraaf to the advancement of the state of the science of mathematics. Issues are published as S. K. Donaldson frequently as necessary. Each issue will contain only one article. IMRP is expected to publish 400± pages in 2006. Yakov Eliashberg Edward Frenkel Articles of at least 50 pages are welcome and all articles are refereed and judged for Emmanuel Hebey correctness, interest, originality, depth, and applicability. Submissions are made by e-mail to Dennis Hejhal [email protected].
    [Show full text]
  • On Hypercomplex Number Systems*
    ON HYPERCOMPLEX NUMBER SYSTEMS* (FIRST PAPER) BY HENRY TABER Introduction. The method invented by Benjamin Peirce t for treating the problem to determine all hypereomplex number systems (or algebras) in a given number of units depends chiefly, first, upon the classification of hypereomplex number sys- tems into idempotent number systems, containing one or more idempotent num- bers, \ and non-idempotent number systems containing no idempotent number ; and, second, upon the regularizaron of idempotent number systems, that is, the classification of each of the units of such a system with respect to one of the idempotent numbers of the system. For the purpose of such classification and regularizaron the following theorems are required : Theorem I.§ In any given hypereomplex number system there is an idem- potent number (that is, a number I + 0 such that I2 = I), or every number of the system is nilpotent. || Theorem H.*rj In any hypereomplex number system containing an idem- potent number I, the units ex, e2, ■■ -, en can be so selected that, with reference ♦Presented to the Society February 27, 1904. Received for publication February 27, 1904, and September 6, 1904. t American Journal of Mathematics, vol. 4 (1881), p. 97. This work, entitled Linear Associative Algebra, was published in lithograph in 1870. For an estimate of Peirce's work and for its relation to the work of Study, Scheffers, and others, see articles by H. E. Hawkes in the American Journal of Mathematics, vol. 24 (1902), p. 87, and these Transactions, vol. 3 (1902), p. 312. The latter paper is referred to below when reference is made to Hawkes' work.
    [Show full text]
  • Pure Spinors to Associative Triples to Zero-Divisors
    Pure Spinors to Associative Triples to Zero-Divisors Frank Dodd (Tony) Smith, Jr. - 2012 Abstract: Both Clifford Algebras and Cayley-Dickson Algebras can be used to construct Physics Models. Clifford and Cayley-Dickson Algebras have in common Real Numbers, Complex Numbers, and Quaternions, but in higher dimensions Clifford and Cayley-Dickson diverge. This paper is an attempt to explore the relationship between higher-dimensional Clifford and Cayley-Dickson Algebras by comparing Projective Pure Spinors of Clifford Algebras with Zero-Divisor structures of Cayley-Dickson Algebras. 1 Pure Spinors to Associative Triples to Zero-Divisors Frank Dodd (Tony) Smith, Jr. - 2012 Robert de Marrais and Guillermo Moreno, pioneers in studying Zero-Divisors, unfortunately have passed (2011 and 2006). Pure Spinors to Associative Triples ..... page 2 Associative Triples to Loops ............... page 3 Loops to Zero-Divisors ....................... page 6 New Phenomena at C32 = T32 ........... page 9 Zero-Divisor Annihilator Geometry .... page 17 Pure Spinors to Associative Triples Pure spinors are those spinors that can be represented by simple exterior wedge products of vectors. For Cl(2n) they can be described in terms of bivectors Spin(2n) and Spin(n) based on the twistor space Spin(2n)/U(n) = Spin(n) x Spin(n). Since (1/2)((1/2)(2n)(2n-1)-n^2) = (1/2)(2n^2-n-n^2) = = (1/2)(n(n-1)) Penrose and Rindler (Spinors and Spacetime v.2) describe Cl(2n) projective pure half-spinors as Spin(n) so that the Cl(2n) full space of pure half-spinors has dimension dim(Spin(n))
    [Show full text]
  • Theory of Trigintaduonion Emanation and Origins of Α and Π
    Theory of Trigintaduonion Emanation and Origins of and Stephen Winters-Hilt Meta Logos Systems Aurora, Colorado Abstract In this paper a specific form of maximal information propagation is explored, from which the origins of , , and fractal reality are revealed. Introduction The new unification approach described here gives a precise derivation for the mysterious physics constant (the fine-structure constant) from the mathematical physics formalism providing maximal information propagation, with being the maximal perturbation amount. Furthermore, the new unification provides that the structure of the space of initial ‘propagation’ (with initial propagation being referred to as ‘emanation’) has a precise derivation, with a unit-norm perturbative limit that leads to an iterative-map-like computed (a limit that is precisely related to the Feigenbaum bifurcation constant and thus fractal). The computed can also, by a maximal information propagation argument, provide a derivation for the mathematical constant . The ideal constructs of planar geometry, and related such via complex analysis, give methods for calculation of to incredibly high precision (trillions of digits), thereby providing an indirect derivation of to similar precision. Propagation in 10 dimensions (chiral, fermionic) and 26 dimensions (bosonic) is indicated [1-3], in agreement with string theory. Furthermore a preliminary result showing a relation between the Feigenbaum bifurcation constant and , consistent with the hypercomplex algebras indicated in the Emanator Theory, suggest an individual object trajectory with 36=10+26 degrees of freedom (indicative of heterotic strings). The overall (any/all chirality) propagation degrees of freedom, 78, are also in agreement with the number of generators in string gauge symmetries [4].
    [Show full text]
  • Truly Hypercomplex Numbers
    TRULY HYPERCOMPLEX NUMBERS: UNIFICATION OF NUMBERS AND VECTORS Redouane BOUHENNACHE (pronounce Redwan Boohennash) Independent Exploration Geophysical Engineer / Geophysicist 14, rue du 1er Novembre, Beni-Guecha Centre, 43019 Wilaya de Mila, Algeria E-mail: [email protected] First written: 21 July 2014 Revised: 17 May 2015 Abstract Since the beginning of the quest of hypercomplex numbers in the late eighteenth century, many hypercomplex number systems have been proposed but none of them succeeded in extending the concept of complex numbers to higher dimensions. This paper provides a definitive solution to this problem by defining the truly hypercomplex numbers of dimension N ≥ 3. The secret lies in the definition of the multiplicative law and its properties. This law is based on spherical and hyperspherical coordinates. These numbers which I call spherical and hyperspherical hypercomplex numbers define Abelian groups over addition and multiplication. Nevertheless, the multiplicative law generally does not distribute over addition, thus the set of these numbers equipped with addition and multiplication does not form a mathematical field. However, such numbers are expected to have a tremendous utility in mathematics and in science in general. Keywords Hypercomplex numbers; Spherical; Hyperspherical; Unification of numbers and vectors Note This paper (or say preprint or e-print) has been submitted, under the title “Spherical and Hyperspherical Hypercomplex Numbers: Merging Numbers and Vectors into Just One Mathematical Entity”, to the following journals: Bulletin of Mathematical Sciences on 08 August 2014, Hypercomplex Numbers in Geometry and Physics (HNGP) on 13 August 2014 and has been accepted for publication on 29 April 2015 in issue No.
    [Show full text]
  • Coset Group Construction of Multidimensional Number Systems
    Symmetry 2014, 6, 578-588; doi:10.3390/sym6030578 OPEN ACCESS symmetry ISSN 2073-8994 www.mdpi.com/journal/symmetry Article Coset Group Construction of Multidimensional Number Systems Horia I. Petrache Department of Physics, Indiana University Purdue University Indianapolis, Indianapolis, IN 46202, USA; E-Mail: [email protected]; Tel.: +1-317-278-6521; Fax: +1-317-274-2393 Received: 16 June 2014 / Accepted: 7 July 2014 / Published: 11 July 2014 Abstract: Extensions of real numbers in more than two dimensions, in particular quaternions and octonions, are finding applications in physics due to the fact that they naturally capture symmetries of physical systems. However, in the conventional mathematical construction of complex and multicomplex numbers multiplication rules are postulated instead of being derived from a general principle. A more transparent and systematic approach is proposed here based on the concept of coset product from group theory. It is shown that extensions of real numbers in two or more dimensions follow naturally from the closure property of finite coset groups adding insight into the utility of multidimensional number systems in describing symmetries in nature. Keywords: complex numbers; quaternions; representations 1. Introduction While the utility of the familiar complex numbers in physics and applied sciences is not questionable, one important question still stands: where do they come from? What are complex numbers fundamentally? This question becomes even more important considering that complex numbers in more than two dimensions such as quaternions and octonions are gaining renewed interest in physics. This manuscript demonstrates that multidimensional numbers systems are representations of small group symmetries. Although measurements in the laboratory produce real numbers, complex numbers in two dimensions are essential elements of mathematical descriptions of physical systems.
    [Show full text]
  • The Cayley-Dickson Construction in ACL2
    The Cayley-Dickson Construction in ACL2 John Cowles Ruben Gamboa University of Wyoming Laramie, WY fcowles,[email protected] The Cayley-Dickson Construction is a generalization of the familiar construction of the complex numbers from pairs of real numbers. The complex numbers can be viewed as two-dimensional vectors equipped with a multiplication. The construction can be used to construct, not only the two-dimensional Complex Numbers, but also the four-dimensional Quaternions and the eight-dimensional Octonions. Each of these vector spaces has a vector multiplication, v1 • v2, that satisfies: 1. Each nonzero vector, v, has a multiplicative inverse v−1. 2. For the Euclidean length of a vector jvj, jv1 • v2j = jv1j · jv2j Real numbers can also be viewed as (one-dimensional) vectors with the above two properties. ACL2(r) is used to explore this question: Given a vector space, equipped with a multiplication, satisfying the Euclidean length condition 2, given above. Make pairs of vectors into “new” vectors with a multiplication. When do the newly constructed vectors also satisfy condition 2? 1 Cayley-Dickson Construction Given a vector space, with vector addition, v1 + v2; vector minus −v; a zero vector~0; scalar multiplica- tion by real number a, a ◦ v; a unit vector~1; and vector multiplication v1 • v2; satisfying the Euclidean length condition jv1 • v2j = jv1j · jv2j (2). 2 2 Define the norm of vector v by kvk = jvj . Since jv1 • v2j = jv1j · jv2j is equivalent to jv1 • v2j = 2 2 jv1j · jv2j , the Euclidean length condition is equivalent to kv1 • v2k = kv1k · kv2k: Recall the dot (or inner) product, of n-dimensional vectors, is defined by (x1;:::;xn) (y1;:::;yn) = x1 · y1 + ··· + xn · yn n = ∑ xi · yi i=1 Then Euclidean length and norm of vector v are given by p jvj = v v kvk = v v: Except for vector multiplication, it is easy to treat ordered pairs of vectors, (v1;v2), as vectors: 1.
    [Show full text]
  • Hypercomplex Numbers and Their Matrix Representations
    Hypercomplex numbers and their matrix representations A short guide for engineers and scientists Herbert E. M¨uller http://herbert-mueller.info/ Abstract Hypercomplex numbers are composite numbers that sometimes allow to simplify computations. In this article, the multiplication table, matrix represen- tation and useful formulas are compiled for eight hypercomplex number systems. 1 Contents 1 Introduction 3 2 Hypercomplex numbers 3 2.1 History and basic properties . 3 2.2 Matrix representations . 4 2.3 Scalar product . 5 2.4 Writing a matrix in a hypercomplex basis . 6 2.5 Interesting formulas . 6 2.6 Real Clifford algebras . 8 ∼ 2.7 Real numbers Cl0;0(R) = R ......................... 10 3 Hypercomplex numbers with 1 generator 10 ∼ 3.1 Bireal numbers Cl1;0(R) = R ⊕ R ...................... 10 ∼ 3.2 Complex numbers Cl0;1(R) = C ....................... 11 4 Hypercomplex numbers with 2 generators 12 ∼ ∼ 4.1 Cockle quaternions Cl2;0(R) = Cl1;1(R) = R(2) . 12 ∼ 4.2 Hamilton quaternions Cl0;2(R) = H ..................... 13 5 Hypercomplex numbers with 3 generators 15 ∼ ∼ 5.1 Hamilton biquaternions Cl3;0(R) = Cl1;2(R) = C(2) . 15 ∼ 5.2 Anonymous-3 Cl2;1(R) = R(2) ⊕ R(2) . 17 ∼ 5.3 Clifford biquaternions Cl0;3 = H ⊕ H .................... 19 6 Hypercomplex numbers with 4 generators 20 ∼ ∼ ∼ 6.1 Space-Time Algebra Cl4;0(R) = Cl1;3(R) = Cl0;4(R) = H(2) . 20 ∼ ∼ 6.2 Anonymous-4 Cl3;1(R) = Cl2;2(R) = R(4) . 21 References 24 A Octave and Matlab demonstration programs 25 A.1 Cockle Quaternions . 25 A.2 Hamilton Bi-Quaternions .
    [Show full text]
  • Geodesics in Hypercomplex Number Systems. Application to Commutative Quaternions
    Dipartimento Energia GEODESICS IN HYPERCOMPLEX NUMBER SYSTEMS. APPLICATION TO COMMUTATIVE QUATERNIONS FRANCESCO CATONI, PAOLO ZAMPETTI ENEA - Dipartimento Energia Centro Ricerche Casaccia, Roma ROBERTO CANNATA ENEA - Funzione Centrals INFO Centro Ricerche Casaccia, Roma LUCIANA BORDONI ENEA - Funzione Centrals Studi Centro Ricerche Casaccia, Roma R eCE#VSQ J foreign SALES PROHIBITED RT/EFtG/97/11 ENTE PER LE NUOVE TECNOLOGIE, L'ENERGIA E L'AMBIENTE Dipartimento Energia GEODESICS IN HYPERCOMPLEX NUMBER SYSTEMS. APPLICATION TO COMMUTATIVE QUATERNIONS FRANCESCO CATONI, PAOLO ZAMPETTI ENEA - Dipartimento Energia Centro Ricerche Casaccia, Roma ROBERTO CANNATA ENEA - Funzione Centrale INFO Centro Ricerche Casaccia, Roma LUCIANA GORDON! ENEA - Funzione Centrale Studi Centro Ricerche Casaccia, Roma RT/ERG/97/11 Testo pervenuto nel giugno 1997 I contenuti tecnico-scientifici del rapporti tecnici dell'ENEA rispecchiano I'opinione degli autori e non necessariamente quella dell'Ente. DISCLAIMER Portions of this document may be illegible electronic image products. Images are produced from the best available original document. RIASSUNTO Geodetiche nei sistemi di numeri ipercomplessi. Appli­ cations ai quaternion! commutativi Alle funzioni di variabili ipercomplesse sono associati dei campi fisici. Seguendo le idee che hanno condotto Einstein alia formulazione della Relativita Generale, si intende verificare la possibility di descrivere il moto di un corpo in un campo gravitazionale, mediante le geodetiche negli spazi ’’deformati ” da trasformazioni funzionali di variabili ipercomplesse che introducono nuove simmetrie dello spazio. II presente lavoro rappresenta la fase preliminare di questo studio pin ampio. Come prima applicazione viene studiato il sistema particolare di numeri, chiamato ”quaternion! commutativi ”, che possono essere considerati come composizione dei sistemi bidimensionali dei numeri complessi e dei numeri iperbolici.
    [Show full text]
  • Similarity and Consimilarity of Elements in Real Cayley-Dickson Algebras
    SIMILARITY AND CONSIMILARITY OF ELEMENTS IN REAL CAYLEY-DICKSON ALGEBRAS Yongge Tian Department of Mathematics and Statistics Queen’s University Kingston, Ontario, Canada K7L 3N6 e-mail:[email protected] October 30, 2018 Abstract. Similarity and consimilarity of elements in the real quaternion, octonion, and sedenion algebras, as well as in the general real Cayley-Dickson algebras are considered by solving the two fundamental equations ax = xb and ax = xb in these algebras. Some consequences are also presented. AMS mathematics subject classifications: 17A05, 17A35. Key words: quaternions, octonions, sedenions, Cayley-Dickson algebras, equations, similarity, consim- ilarity. 1. Introduction We consider in the article how to establish the concepts of similarity and consimilarity for ele- n arXiv:math-ph/0003031v1 26 Mar 2000 ments in the real quaternion, octonion and sedenion algebras, as well as in the 2 -dimensional real Cayley-Dickson algebras. This consideration is motivated by some recent work on eigen- values and eigenvectors, as well as similarity of matrices over the real quaternion and octonion algebras(see [5] and [18]). In order to establish a set of complete theory on eigenvalues and eigenvectors, as well as similarity of matrices over the quaternion, octonion, sedenion alge- bras, as well as the general real Cayley-Dickson algebras, one must first consider a basic problem— how to characterize similarity of elements in these algebras, which leads us to the work in this article. Throughout , , and denote the real quaternion, octonion, and sedenion algebras, H O S n respectively; n denotes the 2 -dimensional real Cayley-Dickson algebra, and denote A R C the real and complex number fields, respectively.
    [Show full text]
  • Algebraic Foundations of Split Hypercomplex Nonlinear Adaptive
    Mathematical Methods in the Research Article Applied Sciences Received XXXX (www.interscience.wiley.com) DOI: 10.1002/sim.0000 MOS subject classification: 60G35; 15A66 Algebraic foundations of split hypercomplex nonlinear adaptive filtering E. Hitzer∗ A split hypercomplex learning algorithm for the training of nonlinear finite impulse response adaptive filters for the processing of hypercomplex signals of any dimension is proposed. The derivation strictly takes into account the laws of hypercomplex algebra and hypercomplex calculus, some of which have been neglected in existing learning approaches (e.g. for quaternions). Already in the case of quaternions we can predict improvements in performance of hypercomplex processes. The convergence of the proposed algorithms is rigorously analyzed. Copyright c 2011 John Wiley & Sons, Ltd. Keywords: Quaternionic adaptive filtering, Hypercomplex adaptive filtering, Nonlinear adaptive filtering, Hypercomplex Multilayer Perceptron, Clifford geometric algebra 1. Introduction Split quaternion nonlinear adaptive filtering has recently been treated by [23], who showed its superior performance for Saito’s Chaotic Signal and for wind forecasting. The quaternionic methods constitute a generalization of complex valued adaptive filters, treated in detail in [19]. A method of quaternionic least mean square algorithm for adaptive quaternionic has previously been developed in [22]. Additionally, [24] successfully proposes the usage of local analytic fully quaternionic functions in the Quaternion Nonlinear Gradient Descent (QNGD). Yet the unconditioned use of analytic fully quaternionic activation functions in neural networks faces problems with poles due to the Liouville theorem [3]. The quaternion algebra of Hamilton is a special case of the higher dimensional Clifford algebras [10]. The problem with poles in nonlinear analytic functions does not generally occur for hypercomplex activation functions in Clifford algebras, where the Dirac and the Cauchy-Riemann operators are not elliptic [21], as shown, e.g., for hyperbolic numbers in [17].
    [Show full text]