0% found this document useful (0 votes)
39 views

Mathematical Foundations of Computer Science Lecture Outline

This document provides an outline of lecture material on mathematical foundations of computer science. It includes examples and solutions related to equivalence relations, equivalence classes, representing relations using directed graphs, and the probabilistic method. Key topics covered are determining the number of reflexive relations on a set, showing that congruent modulo m is an equivalence relation, and using the probabilistic method to prove there is a tournament graph with at least n!/2n-1 Hamiltonian paths.

Uploaded by

Chenyang Fang
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
0% found this document useful (0 votes)
39 views

Mathematical Foundations of Computer Science Lecture Outline

This document provides an outline of lecture material on mathematical foundations of computer science. It includes examples and solutions related to equivalence relations, equivalence classes, representing relations using directed graphs, and the probabilistic method. Key topics covered are determining the number of reflexive relations on a set, showing that congruent modulo m is an equivalence relation, and using the probabilistic method to prove there is a tournament graph with at least n!/2n-1 Hamiltonian paths.

Uploaded by

Chenyang Fang
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/ 5

Mathematical Foundations of Computer Science

Lecture Outline
November 27, 2018

Example. How many reflexive relations are there on a set A of size n?

Solution. We know that R ⊆ A × A. The procedure of constructing a reflexive relation


R is as follows:
Step 1: From A × A, include in R all ordered pairs of the form (a, a).
Step 2: For every ordered pair in A × A of the form (a, b), where a 6= b, choose
whether to include it in R or not.
There is one way to do Step 1 and 2n(n−1) ways to do Step 2. By the multiplication rule,
the number of reflexive relations on a set n elements is 2n(n−1) .

Equivalence Relations
A relation R on a set A is an equivalence relation if and only if it is reflexive, symmetric
and transitive.

Example Let m be a positive integer. Show that the congruent modulo m relation
R = {(a, b) : a ≡ b (mod m)}
is an equivalence relation on the set of integers.
(If m is a positive integer then integers x and y are congruent modulo m, written as x ≡ y
(mod m), if m|(x − y)).

Solution. To show that R is an equivalence relation we need to show that it is reflexive,


symmetric, and transitive. R is reflexive because a − a = 0, and 0 = m · 0. R is symmetric
because if a ≡ b (mod m), it means that a−b = m·k, for some integer k. Thus b−a = m(−k)
and hence (b, a) ∈ R. To show that R is transitive, suppose that that a ≡ b (mod m) and
b ≡ c (mod m). Thus, for some integers q1 and q2 , we have a−b = m(q1 ) and b−c = m(q2 ).
Adding these two equations, we get a − c = m(q1 + q2 ) and thus a ≡ c (mod m). Hence R
is transitive.

Example. Suppose that R is the relation on the set of strings of English letters such that
a R b if and only if l(a) = l(b), where l(x) is the length of the string x. Is R an equivalence
relation?

Solution. R is reflexive as l(a) = l(a), for any string a, and hence a R a. Next, suppose
that a R b. This means that l(a) = l(b) and hence l(b) = l(a). Thus b R a and hence R is
symmetric. Finally, suppose that a R b and b R c. Thus l(a) = l(b) and l(b) = l(c), which
implies that l(a) = l(c). Hence a R c and R is transitive. Since R is reflexive, symmetric,
and transitive, it is an equivalence relation.
2 Lecture Outline November 27, 2018

Equivalence Classes
Let R be an equivalence relation on a set A and let a ∈ A. The equivalence class of a,
denoted by [a]R 1 , is the set of all elements of A related (by R) to a; that is

[a]R = {x ∈ A | a R x}

If b ∈ [a]R , then b is called the representative of this equivalence class. Any element in a
class can be used as a representative of the class.

Example. Let R be an equivalence relation on a set A. Then the following statements


for elements a, b ∈ A are equivalent

(i) b ∈ [a] (ii) [a] = [b] (iii) [a] ∩ [b] 6= ∅

Solution. We will prove (i) =⇒ (ii), (ii) =⇒ (iii), and (iii) =⇒ (i).
(i) =⇒ (ii): We will prove the claim by showing that when b ∈ [a], [a] ⊆ [b] and [b] ⊆ [a].
Let c be any arbitrary but particular element in [a]. By definition, a R c. Since b ∈ [a], it
means that a R b, which further implies b R a (since R is symmetric). Since R is transitive
and we know that b R a and a R c, we have b R c and thus c ∈ [b]. We have thus proved that
[a] ⊆ [b].
Let d ∈ [b]. By definition, b R d. We also know that a R b. Since R is transitive, a R b
and b R d, we have a R d. Thus, by definition, d ∈ [a]. We have thus proved that [b] ⊆ [a].

(ii) =⇒ (iii): To prove this we just need to show that [a] 6= ∅. Since R is reflexive, we know
that a ∈ [a]. Since [a] = [b] and [a] is non-empty, it follows that [a] ∩ [b] 6= ∅.

(iii) =⇒ (i): Let c ∈ [a] ∩ [b]. Thus a R c and b R c. Since R is symmetric, we have c R b.
Since R is transitive, a R c and c R b, we have a R b. By, definition b ∈ [a].

Example. Let R be an equivalence relation on a set A. Then the set {[a]R | a ∈ A}


is a partition of the set A. Each element of the set is called an equivalence class of R.
Conversely, given a partition {Ai } of the set A, there is an equivalence relation R that has
sets Ai as its equivalence classes.

Solution. Since each Selement a ∈ A is in its own equivalent class [a], each equivalent
class is non-empty and a∈A [a] = A. From the claim in the previous example (example we
did in last class), for any two elements a and b in A, [a] and [b] are either equal or disjoint.
Thus the equivalent classes partition the set A.
We now prove the converse. Let R be the relation on A that contains all possible pairs
(x, y), where x and y belong to the same subset Ai in the partition. We want to show that
R is reflexive, symmetric and transitive. R is reflexive as any element a ∈ A is in the same
subset of the partition as itself. Next suppose that a R b. This means that a and b are
1
The subscript R in [a]R is dropped when the relation in reference is clear from the context.
November 27, 2018 Lecture Outline 3

in the same subset of the partition of A. Thus, we have b R a and hence R is symmetric.
Finally, suppose that a R b and b R c. This means that a and b are in the same subset of
the partition and so are b and c. This means that a and c are in the same subset of the
partition and hence we have a R c. Thus R is transitive.

Example. If an equivalence relation R is defined by the following set partition on A, then


express R as a set of ordered pairs.

A = {3, 4, 1} ∪ {2}

Solution.

R = {(1, 1), (2, 2), (3, 3), (4, 4), (1, 3), (1, 4), (3, 1), (3, 4), (4, 3), (4, 1)}

Representing Relations Using Directed Graphs


A directed graph, or digraph G = (V, E) consists of a set V of vertices and a subset E ⊆ V ×V
of edges or arcs. An edge of the form (u, u) is represented as an arc from u to itself.
A binary relation R on a set A can be represented as a directed graph in which the
vertices represent the elements of A and for every ordered pair (a, b) ∈ R, there is an
edge from vertex a to vertex b. For example, the digraph corresponding to the relation
R = {(1, 2), (1, 3), (2, 1), (2, 2), (2, 4), (3, 2), (4, 3)} on the set {1, 2, 3, 4} is shown below.

1 2

4 3

The directed graph G representing a relation R can be used to determine properties of


the relation R. R is reflexive iff G contains a self-loop at every vertex. R is symmetric iff
for each edge (a, b) (a 6= b) in G, there is also an edge (b, a) in G. R is antisymmetric iff
for any two distinct vertices a, b there are no edges between them or exactly one of (a, b) or
(b, a) is in G. Thus R is antisymmetric iff for any two distinct vertices a and b, both (a, b)
and (b, a) are not present in G. The relation R is transitive iff edge (u, w) always exists
whenever there is an edge (u, v) and (v, w), for some vertex v.

The Probabilistic Method


A tournament graph is a directed graph with exactly one directed edge between any pair
of vertices. Every tournament graph has at least one Hamiltonian path, a path that visits
4 Lecture Outline November 27, 2018

each vertex exactly once (proved in Hw 9h, Q.6). In 1943, Szele used the Probabilistic
Method to show the existence of a tournament graph with a large number of Hamiltonian
paths. Note that there are tournaments in which there is exactly one Hamiltonian path.
For example, the tournament on vertices {1, 2, . . . , n} in which there is a directed edge (i, j)
iff i < j has exactly one Hamiltonian path.

n!
Example. Prove that there is a n-vertex tournament with at least 2n−1
distinct Hamil-
tonian paths.

Solution. Let G = (n, 1/2) be a n-vertex tournament graph, in which an edge between
any two vertices u and v is directed towards u with probability 12 and towards v with
probability 21 . Let X denote the total number of Hamiltonian paths in G and let Xσ be
an indicator random variable thatP is 1, iff a permutation σ of the vertices in G yields a
Hamiltonian path. Clearly, X = σ Xσ . Applying the Linearity of Expectation, we get
X
E[X] = E[Xσ ]
σ
X
= Pr[Xσ = 1]
σ
X  1 n−1
=
σ
2
n!
=
2n−1
Since a random orientation of the edges, i.e., a random tournament, yields us the above
number in expectation, there must be an orientation of the edges, i.e., a tournament, in
which the number of Hamiltonian paths is at least n!/2n−1 .
An independent set S in G is a subset of vertices such that no two vertices in S share
an edge. The independence number of a graph G, denoted by α(G) is the size of the largest
independent set in G.

Example. Let n be the number of vertices in G and m be the number of edges, and let
d = 2m
n ≥ 1 be the average degree. Then
n
α(G) ≥
2d
This is a weaker version of the celebrated Turán’s theorem.

Solution. Construct a random subset S of vertices by placing each vertex in S indepen-


dently with probability p (to be determined later). Let X be the random variable denoting
the number of vertices in S and let Y be the random variable denoting the number of edges
whose both endpoints are in S. Let Ye be an indicator random variable that is 1 iff both
endpoints of e are in S. By the Linearity of Expectation we have
X X nd 2
E[X] = np and E[Y ] = E[Ye ] = Pr[Ye = 1] = mp2 = p
e e
2
November 27, 2018 Lecture Outline 5

Note that the quantity X − Y denotes the number of vertices in S minus the number of
edges with both endpoints in S. By the Linearity of Expectation we get
 
nd 2 dp
E[X − Y ] = np − p = np 1 −
2 2

This means that there exists a set S such that the number of vertices in S exceeds the
number of edges in S by the above quantity. We now modifyset S by deleting an arbitrary
0 dp
endpoint of each edge. The resulting set S has at least np 1 − 2 vertices left and has
no edges between any of its vertices. We want to maximize |S 0 |, so we set p = 1/d (using
d ≥ 1), giving us |S 0 | = 2d
n
.

You might also like