v2009.01.01 - Convex Optimization

v2009.01.01 - Convex Optimization v2009.01.01 - Convex Optimization

convexoptimization.com
from convexoptimization.com More from this publisher
10.03.2015 Views

38 CHAPTER 2. CONVEX GEOMETRY (a) R (b) R 2 (c) R 3 Figure 12: (a) Ellipsoid in R is a line segment whose boundary comprises two points. Intersection of line with ellipsoid in R , (b) in R 2 , (c) in R 3 . Each ellipsoid illustrated has entire boundary constituted by zero-dimensional faces; in fact, by vertices (2.6.1.0.1). Intersection of line with boundary is a point at entry to interior. These same facts hold in higher dimension. 2.1.7 classical boundary (confer2.6.1.3) Boundary of a set C is the closure of C less its interior presumed nonempty; [48,1.1] which follows from the fact ∂ C = C \ int C (14) int C = C (15) assuming nonempty interior. 2.7 One implication is: an open set has a boundary defined although not contained in the set; interior of an open set equals the set itself. 2.7 Otherwise, for x∈ R n as in (11), [222,2.1-2.3] the empty set is both open and closed. int{x} = ∅ = ∅

2.1. CONVEX SET 39 2.1.7.1 Line intersection with boundary A line can intersect the boundary of a convex set in any dimension at a point demarcating the line’s entry to the set interior. On one side of that entry-point along the line is the exterior of the set, on the other side is the set interior. In other words, starting from any point of a convex set, a move toward the interior is an immediate entry into the interior. [25,II.2] When a line intersects the interior of a convex body in any dimension, the boundary appears to the line to be as thin as a point. This is intuitively plausible because, for example, a line intersects the boundary of the ellipsoids in Figure 12 at a point in R , R 2 , and R 3 . Such thinness is a remarkable fact when pondering visualization of convex polyhedra (2.12,5.14.3) in four dimensions, for example, having boundaries constructed from other three-dimensional convex polyhedra called faces. We formally define face in (2.6). For now, we observe the boundary of a convex body to be entirely constituted by all its faces of dimension lower than the body itself. For example: The ellipsoids in Figure 12 have boundaries composed only of zero-dimensional faces. The two-dimensional slab in Figure 11 is a polyhedron having one-dimensional faces making its boundary. The three-dimensional bounded polyhedron in Figure 16 has zero-, one-, and two-dimensional polygonal faces constituting its boundary. 2.1.7.1.1 Example. Intersection of line with boundary in R 6 . The convex cone of positive semidefinite matrices S 3 + (2.9) in the ambient subspace of symmetric matrices S 3 (2.2.2.0.1) is a six-dimensional Euclidean body in isometrically isomorphic R 6 (2.2.1). The boundary of the positive semidefinite cone in this dimension comprises faces having only the dimensions 0, 1, and 3 ; id est, {ρ(ρ+1)/2, ρ=0, 1, 2}. Unique minimum-distance projection PX (E.9) of any point X ∈ S 3 on that cone S 3 + is known in closed form (7.1.2). Given, for example, λ∈int R 3 + and diagonalization (A.5.2) of exterior point ⎡ X = QΛQ T ∈ S 3 , Λ = ∆ ⎣ ⎤ λ 1 0 λ 2 ⎦ (16) 0 −λ 3

2.1. CONVEX SET 39<br />

2.1.7.1 Line intersection with boundary<br />

A line can intersect the boundary of a convex set in any dimension at a<br />

point demarcating the line’s entry to the set interior. On one side of that<br />

entry-point along the line is the exterior of the set, on the other side is the<br />

set interior. In other words,<br />

starting from any point of a convex set, a move toward the interior is<br />

an immediate entry into the interior. [25,II.2]<br />

When a line intersects the interior of a convex body in any dimension, the<br />

boundary appears to the line to be as thin as a point. This is intuitively<br />

plausible because, for example, a line intersects the boundary of the ellipsoids<br />

in Figure 12 at a point in R , R 2 , and R 3 . Such thinness is a remarkable<br />

fact when pondering visualization of convex polyhedra (2.12,5.14.3) in<br />

four dimensions, for example, having boundaries constructed from other<br />

three-dimensional convex polyhedra called faces.<br />

We formally define face in (2.6). For now, we observe the boundary<br />

of a convex body to be entirely constituted by all its faces of dimension<br />

lower than the body itself. For example: The ellipsoids in Figure 12 have<br />

boundaries composed only of zero-dimensional faces. The two-dimensional<br />

slab in Figure 11 is a polyhedron having one-dimensional faces making its<br />

boundary. The three-dimensional bounded polyhedron in Figure 16 has<br />

zero-, one-, and two-dimensional polygonal faces constituting its boundary.<br />

2.1.7.1.1 Example. Intersection of line with boundary in R 6 .<br />

The convex cone of positive semidefinite matrices S 3 + (2.9) in the ambient<br />

subspace of symmetric matrices S 3 (2.2.2.0.1) is a six-dimensional Euclidean<br />

body in isometrically isomorphic R 6 (2.2.1). The boundary of the<br />

positive semidefinite cone in this dimension comprises faces having only the<br />

dimensions 0, 1, and 3 ; id est, {ρ(ρ+1)/2, ρ=0, 1, 2}.<br />

Unique minimum-distance projection PX (E.9) of any point X ∈ S 3 on<br />

that cone S 3 + is known in closed form (7.1.2). Given, for example, λ∈int R 3 +<br />

and diagonalization (A.5.2) of exterior point<br />

⎡<br />

X = QΛQ T ∈ S 3 , Λ =<br />

∆ ⎣<br />

⎤<br />

λ 1 0<br />

λ 2<br />

⎦ (16)<br />

0 −λ 3

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

Saved successfully!

Ooh no, something went wrong!