Next: 1.3.1 Function spaces, norms,
Up: 1. Introduction
Previous: 1.2.2.2 Second-order conditions
Contents
Index
1.3 Preview of infinite-dimensional
optimization
In Section 1.2 we considered the problem
of minimizing a function
. Now, instead of
we want to allow a general vector space
, and in fact we are interested in
the case when this vector space
is infinite-dimensional. Specifically,
will itself be a space of functions. Let us denote
a generic function in
by
, reserving the letter
for the argument of
. (This
will typically be a scalar, and has
no relation with
from the previous section.)
The function to be minimized
is a real-valued function
on
, which we now denote by
. Since
is a function on a space of
functions, it is called a functional. To summarize, we are minimizing
a functional
.
Unlike in the case of
,
there does not exist a ``universal" function space. Many different choices
for
are possible, and specifying the desired space
is
part of the problem
formulation. Another issue is that in order
to define local minima of
over
, we
need to specify what it means for two functions in
to be close to each other.
Recall that in the definition of a local minimum in Section 1.2, a ball of radius
with respect to the standard Euclidean norm on
was
used to define the notion of closeness. In the present case we will again
employ
-balls, but we need to specify which norm we are going to use.
While in
all norms are equivalent (i.e., are within a constant multiple of one
another), in function spaces different choices of a norm lead to drastically
different notions of closeness. Thus, the first thing we need to do is become
more familiar with function spaces and norms on them.
Subsections
Next: 1.3.1 Function spaces, norms,
Up: 1. Introduction
Previous: 1.2.2.2 Second-order conditions
Contents
Index
Daniel
2010-12-20