Suppose we want to maximise (or minimise) the linear function on a set
. Suppose
is defined as the solution of the system of three linear inequalities:
In general, the solution set of these inequalities will be a triangular sea of and
:
The points
are the extreme points on
. We want to prove that
is maximised at
.
Theorem
Suppose that is a closed triangular sea of
and
. Then
attains its maximum at an extreme point of
.
Proof
Proper diagrams to follow…
Suppose is attained at
. This we may assume because
is compact and
is continuous. Now
(*)
Suppose without loss of generality that . Then maximising
is equivalent to maximising
, the
-intercept of the family of lines
of the form (*), with slope
subject to the condition that
.
Assume without loss of generality that . If
, then we can move around in any sense so that
is not a maximum as we can increase the
-intercept of any slope
line:
If
on the line
, then there contains an open ball in
containing
. Hence we can take a point
on the line
with a greater
-intercept.
Hence is on the boundary of
,
. Assume without loss of generality that
is on the line segment
but not equal to
. Let
be the slope of the line segment
(again assume w.l.o.g. that
). If
:
Then the line going through
has a greater
-intercept so
can’t be on the interior of
. Similarly, if
, then the line going through
has a greater
-intercept so
can’t be on the interior
neither.
If :
In this case
is constant on the line segment
so in particular
and hence the maximum is on an extreme point
Leave a comment
Comments feed for this article