10.03.2015 Views

v2009.01.01 - Convex Optimization

v2009.01.01 - Convex Optimization

v2009.01.01 - Convex Optimization

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.

408 CHAPTER 5. EUCLIDEAN DISTANCE MATRIX<br />

0<br />

Figure 107: Elliptope E 2 in isometrically isomorphic R 3 is a line segment<br />

illustrated interior to positive semidefinite cone S 2 + (Figure 37).<br />

5.9.1 Geometric arguments<br />

5.9.1.0.1 Definition. Elliptope: [205] [202,2.3] [96,31.5]<br />

a unique bounded immutable convex Euclidean body in S n ; intersection of<br />

positive semidefinite cone S n + with that set of n hyperplanes defined by unity<br />

main diagonal;<br />

E n ∆ = S n + ∩ {Φ∈ S n | δ(Φ)=1} (981)<br />

a.k.a, the set of all correlation matrices of dimension<br />

dim E n = n(n −1)/2 in R n(n+1)/2 (982)<br />

An elliptope E n is not a polyhedron, in general, but has some polyhedral<br />

faces and an infinity of vertices. 5.42 Of those, 2 n−1 vertices (some extreme<br />

points of the elliptope) are extreme directions yy T of the positive semidefinite<br />

cone, where the entries of vector y ∈ R n belong to {±1} and exercise every<br />

combination. Each of the remaining vertices has rank belonging to the set<br />

{k>0 | k(k + 1)/2 ≤ n}. Each and every face of an elliptope is exposed.<br />

△<br />

5.42 Laurent defines vertex distinctly from the sense herein (2.6.1.0.1); she defines vertex<br />

as a point with full-dimensional (nonempty interior) normal cone (E.10.3.2.1). Her<br />

definition excludes point C in Figure 27, for example.

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

Saved successfully!

Ooh no, something went wrong!