0% found this document useful (0 votes)
7 views4 pages

Understanding Regular Languages and DFAs

Uploaded by

clairezhaolj
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)
7 views4 pages

Understanding Regular Languages and DFAs

Uploaded by

clairezhaolj
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

FINAL REVIEW

RegularLanguages
PEF Q ot states regular if a DFA M St M A
alphapet Ʃ the problemencoded bylang A can
initial state be solved by a DFA
subset F of acceptingstates
transitionfunction 8 Qx Ʃ Q
LCM set of strings w
that M accepts
Distinguishable LowerBound on States
twostrings U v are distinguishable wrt A find a set of k pairwisedistinguishablestrings
if there is string w st only one of K number of states
V W and v w is in A
then 8 90 0
8 190 V
NFA Determinization NFA DFA
generalization of DFA convertanyNFA into DFA
transitionfunction Δ Qx
Ʃ 29 input a b
mapsto a set of states 903
acceptsstring if at
least one 5
execution leads to accepting state
acceptonlyregularlanguages can have exponentialblow up in size
E NFAstate can change w oinput

RegularLanguages RegularExpressions
operations on RegularLanguages
A is regular if
closedunderoperation theresult of
the operation is regular if both languages
are regular
2,8T 5ththat I
3 E NFAM suchthat L M A
prove construct DFA M such that 4 regex r such that h r A
LCM A operation A2
regexoperators
closedunder intersection union romp 1 o or morerepetitions
use productconstruction Q Q Q2 Feltre 2 concert
3 U union
ProvingNon regularity
prove no DFA can accept provenofinite
number of statessufficeto acceptA
supposeinfiniteset of pairwise dist strings with
assumeA is regular K pairwise dist strings
and k numberof strings number of states
is cannot be infinite contradiction
TM Hatting Decidable
finite set of States Q w go go 9r if for every input w TM M is finite term
tapealphabet F w blank symbol_ guaranteed to not enter finite loops
input alphabet Ʃ decidablelanguage A if there ishaltingTM
transitionfunction 8 Q t Q xp yp A hats on9 a 9r
such that L M
Sgo g 8 L R decidable solvable

tangentingghete
trite Trovent goin everyregularlanguage is decidable

ClosureProperties
NTM decidablelanguages are closed under
transition function Δ Qxt 2 ext x L R intersection union comp
Visualize as execution tree either so or 8 prove there existshaltingTM st
doesnot change decidability M accepts inputwhen both M and Ma do
simulate M and Mz
simulatingNTMby PTM causes exponentia
slowdown

Problems DFAS
membership ADFA CM WS M is a DFA and M accepts w 9th w
decidable
1 checkif inputM encodes a DFA
2 exelute M on w and check ifaccepts
emptiness Epf 4M M is a DFA and LCM is empty
decidable
equivalence EQ p M M's M and M are DFAs and LCM 2114
decidable
checkif LCM Arlen empty and
M n LCM empty
take product with negation

Problems Turing Machines

ATM caws M is a TMand M accepts w


needto simulate M on w w universal TM U
if M doesnot halt on w then U does not halt
is decidable
cannotconclude U is halting cannotconclude Any
HALT µ MW I M is aTM and Mhalts on w
undecidable
when M is aTM there is noknownbound on howmanysteps the executionneeds
ATM Erm EQTM HALTTM
all undecidable
Recognizable classify a Problem
if there is TM M St LCM A decidable find haltingTM for it
TM can check if input is [Link] recognizable find TM that acceptsit
maynot terminate
closed under union
M may not terminate ProblemReduction
fix simulate parallelism by alternatingsteps undecidable show
on the two tapes Atmreducesto it
constructTMfor
Theorem if A and NA are is decidable
Attyusing
recognizable Problemmy as subroutine
decidable recognizable
ATM Mins M accepts w
I It s Interightaberdecidable _GivenMin construct in stm
exactlywhen
4 Ep Mx
Classp class up Unrecognizable proof by contrad
show nA to be recognizable
if 1 M decidesA LCM A it NTMaccepts A then both are decidable
2 timecomplexity is o nk in poly time
solvable in poly time if deterministicV st
wis in A iff V accepts CW CS
nondeterministic can guesspossible executions
Boolean Formulas
NP complete CONP
Yoshikiheteres i n D Aisin ND iff nA is in NP
truthassignmentp p Oort 2 A is NP hard B pA B in NP
valid hard UNSAT
SAT 4 63 is satisfiable 1 SAThard INVALID
in NP INVALID in NP
show polytimereduction
CNF AND of clauses ORs
30nF each clause has only 3 literals [Link] ffffltibPutw
ClassPSPACE class NPSPACE NPSPACE PSPACE
if TM M acceptsA acceptedbynondeterministicTM ofspaceO nk
spacetomplexity is o nk prove deterministic non polyspace algo
use guesses and LBR vertex gagr
p PSPACE
QBF Booleanformulas w quantifiers LengthBoundedReachability
TQBF is NP hard PSPACE iomplete is there path set of length K
Recursion LBR S V K 2 and LBR V t K 2
Markovchain foreachsubset T of Q
if T doesnot contain anystate in F 19a
M Q Do T thenif LBR 90 T 2n 1 then reject
TransitionmatrixTfgq probabilityq79 spacecomplexity O na
Markovianprocess Xo Xi X2 if
t dependsonly on Xt 1
Marginalprobability Dtime state

BE it
Transient Recurrent SCCs
willbeabandoned will return every pair u v and usu
allstatesnot in P Xt 9 Xo g 1 maximal
BSCC BottomSCC BSC novertex in U hasoutgoingedge
iff belongstoBSOC
in 1 SCCeitherallrecurrent property everyvertexbelongsto 1 80
or all transient
Reducible
Irreducible Periodicity
possibletimeswhen the process is again in
otherwise hasonly1 SCC state with positiveprobability
if irreducible all period GCD of visit timesV19
statesare recurrent
aperiodic period 1 self loop
sameSCC same period
SteadyStateDistribution LimitingDistribution
if Ds T Ds
Diim116 D1D2
focuson irreducible 1 SCLMC
steadystatedistr is unique maynotalwaysexist in periodicirreducible MC

Ds di dads irreducible exists and unique


diedTat aperiodicand irreducible exists and equals Ds
data Anthi

You might also like