Scribd
Upload
67 views8 pages

Lec 11

This document provides an introduction to convex sets and their properties. [1] A convex set is defined as a set where for any two points within the set, the entire line segment between those points is also contained within the set. [2] Convex combinations are linear combinations of points within a convex set where the coefficients are between 0 and 1 and sum to 1. [3] The convex hull of a set is the smallest convex set that contains the original set, and is formed by taking all possible convex combinations of points within the set.

Uploaded by

Ahmad Dawara
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
67 views8 pages

Lec 11

This document provides an introduction to convex sets and their properties. [1] A convex set is defined as a set where for any two points within the set, the entire line segment between those points is also contained within the set. [2] Convex combinations are linear combinations of points within a convex set where the coefficients are between 0 and 1 and sum to 1. [3] The convex hull of a set is the smallest convex set that contains the original set, and is formed by taking all possible convex combinations of points within the set.

Uploaded by

Ahmad Dawara
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

Applied Optimization for Wireless, Machine Learning, Big Data

Prof. Aditya K. Jagannatham


Department of Electrical Engineering
Indian Institute of Technology, Kanpur

Lecture – 11
Introduction to Convex Sets and Properties

Hello. Welcome to another module in this massive open online course. So, let us discuss
the next topic which is the convex sets.

(Refer Slide Time: 00:36)

One of the important concepts to understand in convex optimization is a convex set. So,
to define a convex set, let us start with the following setup. Consider two n-dimensional
points x1 and x2 . This means that these x1 and x2 are vectors in n-dimensional space. So

there will be a line segment that is joining x1 and x2 .


(Refer Slide Time: 04:39)

Let us say that x is the expression of this line segment which is nothing but a linear
combination.

 x1  1    x2  x

Where  is the weight and 0    1 . The important aspect is the values outside  0,1 are

not allowed to be equal to the  .

So for instance let us take a look at the following points generated by this combination.

If   1, then x  x1.
If   0, then x  x2 .
1 x x
If   , then x  1 2 .
2 2

It is clear from the above observations that for the midpoint of the line segment between
1
x1 and x2 , the value of  must be equal to .
2
(Refer Slide Time: 05:13)

So,  x1  1    x2 traces line segment between x1 and x2 and such a linear combination

is termed as a convex combination. This denotes a point on the line segment between x1

and x2 for any 0    1 .

(Refer Slide Time: 06:11)

Therefore, if any two points x1 and x2 belong to a set S and also the entire line segment

drawn between x1 and x2 belongs to S then such set of points x1 and x2 is known as a
convex set. So the mathematical way of stating a convex set is as follows.
For any points x1 , x2  S ; for all 0    1

 x1  1    x2  S

(Refer Slide Time: 08:51)

Let us take a simple example to illustrate this. For instance, it can be seen that an ellipse
is a convex set. So choose any two points. So the entire line segment joining these two
points lies in the set. So, for any two points, the entire line segment is contained in set
and hence the ellipse is a convex set.

(Refer Slide Time: 10:02)


Similarly, in a hexagon, all the points on the line segment drawn between the pairs of
points which are contained in the hexagon, lies in the hexagon and hence Hexagon is
also a convex set.

On the other hand, consider a region S shown in the above image. It is clear that there are
few pair of points contained in S, which does not satisfy the criteria of a convex set. That
is, the line segment drawn between few pairs of point contained in the set S is not
entirely inside the region. Therefore it implies that the above shown set S is not a convex
set.

(Refer Slide Time: 12:40)

Another important concept associated to the convex set is the convex combination. To
define a convex combination, let us recall the linear combination of x1 and x2 which was

used to define the convex set. Therefore the linear combination  x1  1    x2  S can

be equivalently represented as follows.

 x1  1    x2  1 x1  2 x2

Such that

1  2  1

Also that 1 and  2 satisfies the property of a convex set, therefore


0  1 ,2  1

Thus, such combinations of 1 , 2 are known as the convex combination of x1 and x2 and

the notion of this convex combination is 1 x1  2 x2 .

(Refer Slide Time: 14:49)

Therefore, the generalized notion of convex combination of that includes k points


x1 , x2 , , xk in n dimensional space, is

1 x1  2 x2   k xk

where 1 ,2 , ,k are defined as 1  2   k  1 for each i  R, i  0 .


(Refer Slide Time: 16:32)

Another concept associated to the convex set is the convex hull of a set. The convex hull
is the set of all of the convex combination of points in S.

Now naturally observe that for any convex set S, the convex hull is the set itself; because
if S is a convex set; then it already contains all the convex combinations of the points in
S. So, for a convex set, for instance we saw yesterday that the hexagon is a convex set,
correct.

(Refer Slide Time: 18:36)

For a non-convex set.


In case of non-convex set, consider the non-convex set shown in the above image. In
such type of sets, convex hull is the set which contains the whole non-convex set as well
as all the possible points which lies on the different line segments drawn between any
two points of the non-convex set. Therefore it can be said that the convex hull simply
fills this region around the non-convex set to make it a convex set.

So, let us look at other aspects in the subsequent modules.

You might also like