Theory of Computation

Lecture #4

Sarmad Abbasi

Virtual University

Sarmad Abbasi (Virtual University) Theory of Computation 1 / 26

Lecture 4: Overview

Quickly recall the definition of a Turing machine.

Look at some more examples of Turing machines.
Define the concept of a configuration.
Define what the language accepted by a TM.

Sarmad Abbasi (Virtual University) Theory of Computation 2 / 26

Designing TMs
Last time we looked at the language
n
A = {02 : n ≥ 0}
and we designed a TM for it.

Sarmad Abbasi (Virtual University) Theory of Computation 3 / 26

Designing TMs

Now let us look at another language B

B = {w#w : w ∈ {0, 1}∗ }.

In order to design a TM we must understand the language first.
Let us look at some examples of strings that are in B

0110#0110, 0#0, #, 000#000.

Sarmad Abbasi (Virtual University) Theory of Computation 4 / 26

Designing TMs

Let us look at some examples of strings that are not in B

1010#1100, 0110, #0, 1#, 000#000#, 0#0#0, .

Sarmad Abbasi (Virtual University) Theory of Computation 5 / 26

Designing TMs

The machine will do the following:

1 In the first pass the machine will make sure that the input is of the
form
w1 #w2
where w1 , w2 ∈ {0, 1}∗ .
2 Then the machine will repeatedly scan the input and match each
character of w1 with one character of w2 .

Sarmad Abbasi (Virtual University) Theory of Computation 6 / 26

Designing TMs

We define our TM M
1 Q = {q1 , . . . , q14 , qa , qr }
2 Σ = {0, 1, #}
3 Γ = {0, 1, #, x, }
4 q1 , qa , qr the start, accept and reject states respectively.
5 δ is given by the state diagram.
In this diagram we have not shown qr . All transitions not defined here
go to qr .

Sarmad Abbasi (Virtual University) Theory of Computation 7 / 26

Designing TMs

Transition diagram of the Turing Machine.

Sarmad Abbasi (Virtual University) Theory of Computation 8 / 26

Designing TMs

Let us look at the computation of this TM on the input

011#011.
As you can see in the first phase the machine has converted the input
into

11#011.
Now it is looking for a 0 in the second string. When it finds a 0 it
crosses it and starts all over again.
Next it crosses of the 1 and is looking for a 1 in the second string.......

x1#x11
and is looking for a 1.

Sarmad Abbasi (Virtual University) Theory of Computation 9 / 26

Designing TMs

It finds a 1 after the # and crosses it.

x1#xx1
Now it goes back and similarly matches the last 1 to finally reach

xx#xxx
Then it goes over the tape once and realizes that all things have been
matched so it accepts this input.

Sarmad Abbasi (Virtual University) Theory of Computation 10 / 26

Designing TMs

Put Animation here.

Sarmad Abbasi (Virtual University) Theory of Computation 11 / 26

Designing TMs

Let us look at the another language:

C = {ai bj c k : k = i × j i, j, k ≥ 1}.
Firstly all the strings in this language have a’s then b’s then c’s.
Some strings in the language are:

abc, aabbcccc, aaabbcccccc

Some strings not in the langauge are

ab, aabb, aabbccc, ba, cab,

Sarmad Abbasi (Virtual University) Theory of Computation 12 / 26

Designing TMs

Lets give an informal description of a TM that recognizes C.

On input w
1 Scan the input and make sure that w is of the form a∗ b∗ c ∗ if it is
not reject.
2 Cross off an a and scan to the right till a b occurs. Now, shuttle
between b’s and c’s crossing one b and one c till all b’s are
crossed off. If you run out of c’s reject.
3 If all a’s are crossed off. Check if all c’s are also crossed off. In
that case accept.
4 If there are more a0 s left. Restore all b’s and repeat.

Sarmad Abbasi (Virtual University) Theory of Computation 13 / 26

Designing TMs
On input
aabbbcccccc
We will first check if the string is of the form a∗ b∗ c ∗ and it is.
So in the next state we cross off an a

x 0 abbbcccccc.
Since there are three b’s we shuttle between b’s can c’s crossing them
off.
To get to

x 0 axxxxxxccc
Next there is still an a so we restore all the b’s to get

x 0 abbbxxxccc

Sarmad Abbasi (Virtual University) Theory of Computation 14 / 26

Designing TMs

Now, we cross of another a to get

x 0 xbbbxxxccc.
Now, shuttle between b’s and c’s crossing them off to get

x 0 xxxxxxxxxx
.
Now check if there are any a’s left. No so we accept.

Sarmad Abbasi (Virtual University) Theory of Computation 15 / 26

Designing TMs

One question.
Why is the left most x and x 0 .
The answer is that this way the machine will know that it has reached
the left end of the tape.

Sarmad Abbasi (Virtual University) Theory of Computation 16 / 26

Designing TMs

Why will this work.

Note for each a we are crossing off as many c’s as there are b’s. So if
the number of a’s is i and the number of b’s is j then each time

c’s are being crossed off.

So the total number of c’s crossed off is i × j.
Thus if the number of c’s is equal to

i ×j

we accept.

Sarmad Abbasi (Virtual University) Theory of Computation 17 / 26

Designing TMs

This is a high level description of a TM. It is your home work to convert

this high level description to a formal description.
In your home work you have to specify 7 things. The set of states, the
start, accept and reject state, the input alphabet, the work alphabet
and a transition function (for which you can draw a state diagram).

Sarmad Abbasi (Virtual University) Theory of Computation 18 / 26

Designing TMs

Let us look at another example given in the book.

The language

E = {#x1 #x2 # . . . #xl : each xi ∈ {0, 1}∗ and xi 6= xj for each i 6= j}

This problem is called element distinctness problem.

We want to make sure that all the strings x1 , . . . , xl are distinct.

Sarmad Abbasi (Virtual University) Theory of Computation 19 / 26

Designing TMs

Here is a high level description of the Turing machine M4 that accepts

E.
On input w
1 Place a mark on top of the left most tape symbol. If that symbol is
not # reject.
2 Scan right to the next # and place a second mark on top of it. If
no # is encountered and a blank is encountered accept.
3 By zig-zagging compare the two strings to the right of the marked
#. If they are equal reject.
4 Move the rightmost of the two marks to the next # symbol. If no #
is encountered move the left most marker to the next #. If no such
# is available all strings are compared, so accept.
5 Go to state 3

Sarmad Abbasi (Virtual University) Theory of Computation 20 / 26

Designing TMs

Suppose the input is

#0#10#11#000
We start by marking the first # sign. This means the first string has to
be compared will all the strings.

#0#1011#000
It is first compared with the second string.

#0#10#11#000
then

#0#10#11#000

Sarmad Abbasi (Virtual University) Theory of Computation 21 / 26

Designing TMs

then there are no more # signs. So we move onto 10.

#0#10#11#000
then
#0#10#11#000
then

#0#10#11#000
after comparing them we are done.

Sarmad Abbasi (Virtual University) Theory of Computation 22 / 26

Designing TMs

There is a simple algorithm behind this TM.

Let l be the number of strings on the tape.
for i = 1, 2, . . . , l
for j = i + 1, . . . , l
compare xi and xj if they are equal reject.
accept.

Sarmad Abbasi (Virtual University) Theory of Computation 23 / 26

Designing TMs

From now on. We will only give informal description of TM’s. But, you
should be able to convert any informal description to a formal one.
Most interesting TM have several states hundreds even thousands.
However, it will be clear that an informal description can be changed to
a formal one.

Sarmad Abbasi (Virtual University) Theory of Computation 24 / 26

Variants of TMs
Let us look at an example of a multi-tape Turing machine.
A k -tape TM is a 7-tuple:
1 Q is the set of states.
2 Σ
3 Γ
4 q0 , qa , qr .
5 δ is now a function from
Q × Γk → Q × Γk × {L, R}k .

Sarmad Abbasi (Virtual University) Theory of Computation 25 / 26

Variants of TM’s
Multiple head Turing machines.
A k -tape TM is a 7-tuple:
1 Q is the set of states.
2 Σ
3 Γ
4 q0 , qa , qr .
5 δ is now a function from
Q × Γk → Q × Γk × {L, R}k .

Sarmad Abbasi (Virtual University) Theory of Computation 26 / 26