14.11.2012 Views

Conformal Geometric Algebra in Stochastic Optimization Problems ...

Conformal Geometric Algebra in Stochastic Optimization Problems ...

Conformal Geometric Algebra in Stochastic Optimization Problems ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

2 CHAPTER 1. INTRODUCTION<br />

with a simple, i.e. low dimensional, geometric algebra (<strong>in</strong>dicated <strong>in</strong> brackets). Each<br />

of the quoted constructs possesses its own specific notation even though all of them<br />

can be treated <strong>in</strong> a unified manner by means of GA. A similar situation holds for<br />

the 3D-space with its scalar product and the vector cross product: <strong>in</strong> GA a s<strong>in</strong>gle<br />

operation encompasses both of the ‘constructs’. These then arise naturally <strong>in</strong>sofar<br />

as they reflect the symmetric and anti-symmetric part, respectively, of this only<br />

product. And it does so irrespective of the dimension, i.e. it generalizes to algebras<br />

built on spaces with dimension different from three. The portrayed product is of<br />

course the algebra product - the geometric product. By the excess of <strong>in</strong>formation,<br />

it follows that, not least from a sheerly geometrical po<strong>in</strong>t of view, the geometric<br />

product must be reversible if one of its operands is still known. It is hence the<br />

accomplishment of the geometric product that virtually all elements of the algebra<br />

have a multiplicative <strong>in</strong>verse. As the algebra always comprises the vector space it<br />

is built on, it can <strong>in</strong> particular be ‘divided’ by a vector.<br />

Clearly, every task may likewise be accomplished, <strong>in</strong> a way, by means of the standard<br />

vector algebra. It may be permitted to say: ‘this is comparable to a pipe fitter<br />

who <strong>in</strong>cessantly utilizes hammer and screwdriver, although, for example, the pipe<br />

wrench <strong>in</strong> his toolkit beside him would be much more practicable.’ S<strong>in</strong>ce GA can<br />

be regarded as a universal framework, problems should be modeled with it from the<br />

beg<strong>in</strong>n<strong>in</strong>g on. It is then still possible to make the transition to a suitable matrix<br />

representation if a numerical evaluation is due. This is exactly the way chosen <strong>in</strong><br />

this work - no model<strong>in</strong>g option is abandoned before the time.<br />

1.1.1 GA - General Th<strong>in</strong>gs<br />

The term �p,q denotes the geometric algebra over an n-dimensional vector space<br />

�n , n = p + q, equipped with signature (p,q) ∈ �2 . The latter will, however,<br />

become of importance not before the next chapter. The dimension of the algebra<br />

is then 2n . Its elements are termed multivectors, where it is fully sufficient to treat<br />

a multivector as a vector from � 2n<br />

as long as its respective algebra is known. The<br />

components of a multivector comprise n+1 different grades, each with a multiplicity<br />

accord<strong>in</strong>g to the n + 1th row of Pascal’s triangle. By convention, the components<br />

of a multivector are ordered accord<strong>in</strong>g to their grade: one scalar (grade zero), n<br />

vectors (grade one) and the higher grades. Note that this one-to-one corresponds<br />

to the quadruplet representation of the quaternions with basis 1 = (1,0,0,0), i =<br />

(0, 1, 0, 0), j = (0,0,1,0) and k = (0,0,0,1). Hence, impos<strong>in</strong>g the common order<br />

1,i, j,k, multivector k, for <strong>in</strong>stance, correctly reflects the element of grade two.<br />

Moreover, observe that just like for the complex numbers, e.g. z = a + ib, a<br />

number a can be added to a vector ib <strong>in</strong> a mean<strong>in</strong>gful way. The simple relationships<br />

can be looked up <strong>in</strong> many textbooks. Important for this work is as well that for<br />

every GA there exists at least one matrix representation with real square matrices<br />

represent<strong>in</strong>g multivectors. Build<strong>in</strong>g the geometric product simply amounts to the<br />

matrix product. An illustrative example are the Pauli matrices (of size 4 × 4 <strong>in</strong><br />

the real case): the identity matrix holds the scalar component. The Pauli matrices<br />

themselves can be identified with the three vector valued basis elements of the<br />

eight-dimensional Pauli algebra. The existence of the isomorphism between GA and

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!