Scribd
Upload
1 views

Graph Coloring-III

Graph Coloring-3

Uploaded by

Salik Ahmad
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
1 views

Graph Coloring-III

Graph Coloring-3

Uploaded by

Salik Ahmad
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 14

National University of Computer & Emerging Sciences

MT-3001 Graph Theory


Instructor: Miss Urooj
GRAPH COLORING
A graph can have a chromatic number that is much larger than its clique size.
In fact, Jan Mycielski showed that there exist graphs with an arbitrarily large chromatic number yet
have a clique size of 2.
• We often refer to graphs with ω(G) = 2 as triangle-free.
• Mycielski’s proof provided a method for finding a triangle-free graph that requires the desired
number of colors.

1
If we continue this procedure through one more step, we obtain a graph needing 5 colors with a
clique size of 2.
Remark:
Although Mycielski’s Construction should warn you not to rely too heavily on the clique size of a
graph, most real-world applications have a chromatic number close to their clique size.

2
Brooks’ Theorem:

COLORING STRATEGIES:
The bounds above provide starting points for determining the range in which to search for a
proper k-coloring of a graph.
The process for finding a minimum coloring is not trivial, though we will discuss some strategies for
determining the chromatic number of a graph.
Intuitive Idea:
 In our discussion of Brooks’ Theorem, we noted that if every neighbor of a vertex has a
different color, then one additional color would be needed for that vertex.
 This implies that large degree vertices are more likely to increase the value for the
chromatic number of a graph and thus should be assigned a color earlier rather than later in
the process.
 In addition, it is better to look for locations in which colors are forced rather than chosen;
that is, once an initial vertex is given color 1, look for cliques within the graph containing
that vertex as these have very clear restrictions on assigning future colors.

3
4
Remark:
The coloring obtained in Example 6.3 was not unique.
There are many ways to find a proper coloring for the graph; however, every proper coloring
would need at least five colors.
Question:
In terms of the graph model (forming teams) does the solution above seem fair?
Often, we are not only looking for the minimal k-coloring, but also one that adds in a notion of
fairness.
5
Equitable coloring:

By this definition, the final coloring from Example 6.3 is not equitable.
Note that not all graphs have equitable coloring using exactly χ(G) colors.

Below is a summary of the coloring strategies we have discussed so far.

6
7
General Results:
 First, we begin with a basic counting argument relating the number of edges of a graph with
its chromatic number.
 In essence we can create an upper bound based on the number of edges in a graph rather
than the maximum degree.

The next two results find upper bounds on the chromatic number based on structures within the
graph, the first based on the length of the longest path and the second using induced subgraphs.

8
The main reason we need induced subgraphs for coloring problems is that if we took any subgraph
and colored it, we may be missing edges that would indicate two vertices need different colors in
the larger graph.

Perfect Graphs:

9
As we have already seen, having an induced graph isomorphic to C5 would disqualify a graph from
being perfect.

10
Example of Chordal Graph:

11
12
Remark:
It should be noted that in most applications of interval graphs, you are given the intervals and
must form the graph. A much harder problem is determining if an interval representation of a
graph exists and then finding one. But once an interval representation is known, coloring interval
graphs is quite easy by simply coloring the vertices based on when the corresponding interval is
first seen as we sweep from left to right.

13
14

You might also like