0% found this document useful (0 votes)
51 views67 pages

Turing's Thesis: Fall 2006 Costas Busch - RPI 1

Turing's thesis states that any computation that can be performed by mechanical means can be performed by a Turing machine. An algorithm is a Turing machine that solves a given problem. When we say there exists an algorithm for a problem, we mean there exists a Turing machine that executes that algorithm. Variations of the Turing machine model, including those with multiple tapes, semi-infinite tapes, off-line input, nondeterminism, and multiple dimensions all have the same computational power as the standard Turing machine model. This computational equivalence is proven through simulations that map one machine model to another while accepting the same formal languages.

Uploaded by

Raggy
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
51 views67 pages

Turing's Thesis: Fall 2006 Costas Busch - RPI 1

Turing's thesis states that any computation that can be performed by mechanical means can be performed by a Turing machine. An algorithm is a Turing machine that solves a given problem. When we say there exists an algorithm for a problem, we mean there exists a Turing machine that executes that algorithm. Variations of the Turing machine model, including those with multiple tapes, semi-infinite tapes, off-line input, nondeterminism, and multiple dimensions all have the same computational power as the standard Turing machine model. This computational equivalence is proven through simulations that map one machine model to another while accepting the same formal languages.

Uploaded by

Raggy
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 67

Turings Thesis

Turings thesis (1930):


Any computation carried out by mechanical means can be performed by a Turing Machine

Algorithm: An algorithm for a problem is a Turing Machine which solves the problem

The algorithm describes the steps of the mechanical means This is easily translated to computation steps of a Turing machine

When we say: There exists an algorithm

We mean: There exists a Turing Machine that executes the algorithm

Variations of the Turing Machine

The %tandard Model


&nfinite Tape

aababbcac a
Read Write !ead "ontrol #nit $eterministic '(eft or Right)

Variations of the %tandard Model


Turing machines with: * %tay +ption * %emi &nfinite Tape * +ff (ine * Multitape * Multidimensional * ,ondeterministic $ifferent Turing Machine Classes

%ame -ower of two machine classes: both classes accept the same set of languages We will prove: each new class has the same power with %tandard Turing Machine 'accept Turing Recogni.able (anguages)

%ame -ower of two classes means: for any machine

M1

of first class of second class

there is a machine such that:

M2

L( M1 ) = L( M 2 )
and vice versa

Simulation: A techni/ue to prove same power0 %imulate the machine of one class with a machine of the other class %econd "lass %imulation Machine

1irst "lass +riginal Machine

M2 M1
simulates

M1

M1

"onfigurations in the +riginal Machine have corresponding configurations in the %imulation Machine M 2

M1

+riginal Machine:

d0

M1 d1 d n

%imulation Machine:

d1 dn d0 M2

Accepting "onfiguration +riginal Machine:

df

%imulation Machine:

d f

the %imulation Machine and the +riginal Machine accept the same strings

L( M1 ) = L( M 2 )

Turing Machines with %tay +ption

The head can stay in the same position

aababbcac a
(eft2 Right2 %tay (2R2%: possible head moves

3xample:

Time 4

aababbcac a q1
Time 5

b ab abb c ac a q2 q1 a b, S q2

Theorem: %tay +ption machines have the same power with %tandard Turing machines Proof: 40 %tay +ption Machines simulate %tandard Turing machines 50 %tandard Turing machines simulate %tay +ption machines

40 %tay +ption Machines simulate %tandard Turing machines

Trivial: any standard Turing machine is also a %tay +ption machine

50 %tandard Turing machines simulate %tay +ption machines

We need to simulate the stay head option with two head moves2 one left and one right

%tay +ption Machine

q1

a b, S

q2

%imulation in %tandard Machine

q1

a b, L

x x, R

q2

1or every possible tape symbol

1or other transitions nothing changes %tay +ption Machine

q1

a b, L

q2

%imulation in %tandard Machine

q1

a b, L

q2

%imilar for Right moves

example of simulation

q1 a b, S q2

%tay +ption Machine: 4

aaba q1 baba q2

baba q2 baba
q3

%imulation in %tandard Machine: 4

aaba q1

3,$ +1 -R++1

Multiple Trac7 Tape


A useful tric7 to perform more complicated simulations +ne Tape

a b a b b a c d
+ne head +ne symbol ( a, b)

trac7 4 trac7 5

a b a b b a c d q1 a c a b b d c d q2 q1
(b, a ) (c, d ), L

trac7 4 trac7 5

trac7 4 trac7 5

q2

%emi &nfinite Tape


The head extends infinitely only to the right

a b a c

000000000

* &nitial position is the leftmost cell * When the head moves left from the border2 it returns to the same position

Theorem:

%emi &nfinite machines have the same power with %tandard Turing machines

Proof: 40 %tandard Turing machines simulate %emi &nfinite machines 50 %emi &nfinite Machines simulate %tandard Turing machines

40 %tandard Turing machines simulate %emi &nfinite machines:


# a b a c

%tandard Turing Machine


a0 insert special symbol at left of input string b0 Add a self loop to every state
'except states with no outgoing transitions)

#
# # , R

50 %emi &nfinite tape machines simulate %tandard Turing machines: 000000000 %tandard machine 000000000

%emi &nfinite tape machine 000000000 %/uee.e infinity of both directions in one direction

000000000

%tandard machine

a b c d e
reference point

000000000

%emi &nfinite tape machine with two trac7s Right part # d e (eft part

# c b a

000000000

%tandard machine

q1

q2

%emi &nfinite tape machine (eft part


L q1 L q2

Right part
R q1 R q2

%tandard machine

q1

a g, R

q2

%emi &nfinite tape machine Right part (eft part


R q1 L q1

(a, x) ( g , x), R ( x, a ) ( x, g ), L

R q2 L q2

1or all tape symbols

Time 4 000000000

%tandard machine

a b c d e q1
%emi &nfinite tape machine

000000000

Right part # d e (eft part

# c b a
L q1

000000000

Time 5 000000000

%tandard machine

g b c d e q2
%emi &nfinite tape machine

000000000

Right part # d e (eft part

g c b #
L q2

000000000

At the border:

%emi &nfinite tape machine Right part


R q1

(# , # ) (# , # ), R

L q1

(eft part

L q1

(# , # ) (# , # ), R

R q1

%emi &nfinite tape machine Right part # d e (eft part


L q1

# c b g
Time 5

Time 4

000000000

Right part # d e (eft part


R q1

# c b g

000000000

3,$ +1 -R++1

The +ff (ine Machine


&nput 1ile read only 'once)

a b c
&nput string "ontrol #nit
'state machine)

&nput string Appears on input file only

g d e

Tape

read write

Theorem: +ff (ine machines have the same power with %tandard Turing machines Proof: 40 +ff (ine machines simulate %tandard Turing machines 50 %tandard Turing machines simulate +ff (ine machines

40 +ff line machines simulate %tandard Turing Machines +ff line machine: 40 "opy input file to tape 50 "ontinue computation as in %tandard Turing machine

%tandard machine

a b c

+ff line machine

a b c

&nput 1ile

a b c

Tape

40 "opy input file to tape

a b c q1

%tandard machine

+ff line machine

a b c

&nput 1ile

a b c
q1

Tape

50 $o computations as in Turing machine

50 %tandard Turing machines simulate +ff (ine machines:

#se a %tandard machine with a four trac7 tape to 7eep trac7 of the +ff line input file and tape contents

+ff line Machine

a b c d

&nput 1ile

e f g
1our trac7 tape &nput 1ile head position Tape head position

Tape

%tandard Machine

# a b # 0 0 e f 0 1

c d 1 0 g 0

Reference point

'uses special symbol 9 )

# # # #

a 0 e 0

b 0 f 1

c d 1 0 g 0

&nput 1ile head position Tape head position

Repeat for each state transition: 40 Return to reference point 50 1ind current input file symbol 60 1ind current tape symbol 80 Ma7e transition
3,$ +1 -R++1

Multi tape Turing Machines


"ontrol unit
'state machine)

Tape 4

a b c
&nput string

g f e

Tape 5

&nput string appears on Tape 4

Tape 4

Time 4

a b c q1
Tape 4

g f e q1

Tape 5

a g c q2 q1

Time 5

g e d q2

Tape 5

(b, f ) ( g , d ), L, R

q2

Theorem: Multi tape machines have the same power with %tandard Turing machines Proof: 40 Multi tape machines simulate %tandard Turing machines 50 %tandard Turing machines simulate Multi tape machines

40 Multi tape machines simulate %tandard Turing Machines:

Trivial: #se :ust one tape

50 %tandard Turing machines simulate Multi tape machines: %tandard machine: * #ses a multi trac7 tape to simulate the multiple tapes * A tape of the Multi tape machine corresponds to a pair of trac7s

a b c

Multi tape Machine Tape 4 Tape 5

g f e h

%tandard machine with four trac7 tape

a 0 e 0

b 1 f 0

c 0 g h 1 0

Tape 4 head position Tape 5 head position

Reference point

# # # #

a 0 e 0

b 1 f 0

c 0 g h 1 0

Tape 4 head position Tape 5 head position

Repeat for each state transition: 40 Return to reference point 50 1ind current symbol in Tape 4 60 1ind current symbol in Tape 5 80 Ma7e transition

3,$ +1 -R++1

%ame power doesnt imply same speed: n n

L= a b }

%tandard Turing machine: O ( n ) time


2 O ( n ) times ;o bac7 and forth to match the as with the bs

5 tape machine: O ( n) time n 40 "opy b to tape 5 'O(n) steps) 50 "ompare a on tape 4 n and b tape 5 'O(n) steps)

Multidimensional Turing Machines


5 dimensional tape

y c a b

M+V3%: (2R2#2$ #: up $: down

!3A$ -osition: <52 4

Theorem: Multidimensional machines have the same power with %tandard Turing machines Proof: 40 Multidimensional machines simulate %tandard Turing machines 50 %tandard Turing machines simulate Multi $imensional machines

40 Multidimensional machines simulate %tandard Turing machines

Trivial: #se one dimension

50 %tandard Turing machines simulate Multidimensional machines

%tandard machine: * #se a two trac7 tape * %tore symbols in trac7 4 * %tore coordinates in trac7 5

5 dimensional machine

c a b
%tandard Machine

q1
symbols coordinates

c a b 1 # 1 # 2 # 1 # 1 q1

%tandard machine: Repeat for each transition followed in the 5 dimensional machine: 40 #pdate current symbol 50 "ompute coordinates of next position 60 ;o to new position

3,$ +1 -R++1

,ondeterministic Turing Machines a b, L q1 a c, R


q3 "hoice 5

q2

"hoice 4

Allows ,on $eterministic "hoices

Time =

a b c q1 a b, L q1 a c, R
q3

Time 4

q2

b b c q2 c b c
q3
"hoice 5

"hoice 4

&nput string w is accepted if there is a computation:

q0 w x q f y

&nitial configuration

1inal "onfiguration

Any accept state There is a computation:

Theorem: ,ondeterministic machines have the same power with %tandard Turing machines Proof: 40 ,ondeterministic machines simulate %tandard Turing machines 50 %tandard Turing machines simulate ,ondeterministic machines

40 ,ondeterministic Machines simulate %tandard 'deterministic) Turing Machines

Trivial: every deterministic machine is also nondeterministic

50 %tandard 'deterministic) Turing machines simulate ,ondeterministic machines:

$eterministic machine: * #ses a 5 dimensional tape


'which is e/uivalent to 4 dimensional tape)

* %tores all possible computations of the non deterministic machine on the 5 dimensional tape

All possible computation paths &nitial state %tep 4 %tep 5

%tep i re:ect accept


infinite path

%tep i<4

The $eterministic Turing machine simulates all possible computation paths: *simultaneously *step by step *in a breadth first search fashion

,on$eterministic machine

a b, L q1 a c, R # # # #

q2

Time =

a b c q1

q3
$eterministic machine

# # # a b c q1 # # #

# # # # #

current configuration

,on$eterministic machine Time 4

a b, L q1 a c, R

q2

b b c

"hoice 4 "hoice 5

q2 c b c
q3

q3

$eterministic machine

# # # # # # # b b c # # q2 # # c b c # q3 # #

"omputation 4 "omputation 5

$eterministic Turing machine Repeat 1or each configuration in current step of non deterministic machine2 if there are two or more choices: 40 Replicate configuration 50 "hange the state in the replicas #ntil either the input string is accepted or re:ected in all configurations
3,$ +1 -R++1

Remar7: The simulation ta7es in the worst case exponential time compared to the shortest accepting path length of the nondeterministic machine

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy