Chapter 3 Geometry of convex functions - Meboo Publishing ...
Chapter 3 Geometry of convex functions - Meboo Publishing ...
Chapter 3 Geometry of convex functions - Meboo Publishing ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
3.7. GRADIENT 249<br />
f(Y )<br />
[ ∇f(X)<br />
−1<br />
]<br />
∂H −<br />
Figure 75: When a real function f is differentiable at each point in its open<br />
domain, there is an intuitive geometric interpretation <strong>of</strong> function <strong>convex</strong>ity in<br />
terms <strong>of</strong> its gradient ∇f and its epigraph: Drawn is a <strong>convex</strong> quadratic bowl<br />
in R 2 ×R (confer Figure 157, p.668); f(Y )= Y T Y : R 2 → R versus Y on<br />
some open disc in R 2 . Unique strictly supporting hyperplane ∂H − ∈ R 2 × R<br />
(only partially drawn) and its normal vector [ ∇f(X) T −1 ] T at the<br />
particular point <strong>of</strong> support [X T f(X) ] T are illustrated. The interpretation:<br />
At each and every coordinate Y , there is a unique hyperplane containing<br />
[Y T f(Y ) ] T and supporting the epigraph <strong>of</strong> <strong>convex</strong> differentiable f .