3.1 3. P P DTM ( P Turing ) P= DTIME(nk) k 2 2 ( Turing ) • • ( ) 3 4 PATH PATH ={<G,s,t>| G PATH s s t } t ? <G,s,t> s 3. G a 3.2 PATH M=“ 1. 2. P b (a,b) b 4. t PATH M 5 6 RELPRIME RELPRIME ={<x,y> | x y 1 4 1 RELPRIME m 2 (m 3 O(nm) G ) x y ? 3.3 n 1+1+m (relatively prime)} RELPRIME P G M PATH Euclid 7 E 8 a=qb+r (a b gcd(a,b)=gcd(b,r) r 0 ) R r<b E <x,y> <x,y> x y 2 ) 1. <x,y> E ) 2. 1. y=0 2. x y 2 1 x:=x mod y 3. x y 4. x 9 10 3.4 a. 1274 10505 =gcd(740,629) =gcd(22,5) =gcd(629,111) =gcd(5,2) =gcd(111,74) =gcd(2,1) =gcd(74,37) =gcd(1,0)=1 1274 10505 • x/2 ≥ y x gcd(8029,7289)=gcd(7289,740) =gcd(313,22) x<y 3 b a gcd(10505,1274)=gcd(1274,313) 2 ? b. 7289 8029 2 x>y x mod y < y ≤ x/2 x/2<y≤x • x/2 < y x x y =gcd(37,0)=37 3 x y 7289 8029 11 x mod y = x-y < x/2 1 2 12 (Context-Free Language) 3.5 2 3 log x x y log y P (Context-Free Grammar) O(n) ? 4 G=(N,Σ, P, S) N: Σ: S: S N P: P N×(N Σ)* 14 15 (Dynamic Programming) Chomsky 1. Chomsky A→BC 2. A→a a: 1 , A: , B,C: S 3. S→ε 16 17 3.5 ( G 1) 3.5 ( w w=w1…wn 1 n×n (i,j) i≤j wiwi+1…wj 2) 2 18 3.5 ( 19 3) 3.5 ( 4) A→BC B 1 k C 2 A k+1 B C A 2 k (1,n) 20 S 21 3.6 1 w=baba 1 G: 3 R→TR|a 4 T→TR|b 3 4 T S→RT R→TR|a T→TR|b R 2 S→RT 2 T R T→w1 R→w2 T→w3 R→w4 w=w1w2w3w4 =baba 3.5 22 1 2 1 2 T R,T R 3 4 S T 3 R,T 1 S→RT R→TR|a T→TR|b 2 w=w1w2w3w4 =baba 1 2 3 T R,T S R S S T R,T 3 R 4 23 w=w1w2w3w4 =baba 24 S→RT R→TR|a T→TR|b R 4 (R T)→w1w2 S→w2w3 (R T)→w3w4 4 S→w1w2w3 S→w2w3w4 25 1 1 2 3 T R,T S S,R,T R S S T R,T 2 3 4 1 v G S→RT R→TR|a T→TR|b O(n) 2 R 4 n (S w=w1w2w3w4 =baba nv n R n r G T)→w1w2w3w4 r O(n3) n 26 27 (satisfiability problem) (satisfiability problem) ( x=0, y=0, z=0 ) TRUE(=1) FALSE(=0) φ AND( ),OR( ),NOT(¬) 0 0=0 0 0=0 ¬ 0=1 0 1=0 0 1=1 ¬ 1=0 1 0=0 1 0=1 1 1=1 1 1=1 ¬(x y) 1 SAT={<φ> | φ 28 (z x ¬z) 0 1 (satisfiable) } 29 (CNF) 3SAT x ¬x (x1 (x1 ¬x2 ¬x3) (x3 ¬x5 x6) (x3 ¬x6 x4) (x4 x5 x6) x4) ¬x2 ¬x3 3 CNF 3CNF (x1 ¬x2 x4) ¬x3 (x3 ¬x5 x6) (x3 3SAT={<φ> | φ ¬x6) 3CNF } (Conjunctive Normal Form) CNF 30 31 2SAT (x1 ¬x2) (x3 x6) 3.7 (x4 ¬x6) (x5 x6) 2 1. (x y) (x ¬y) (¬x y) (x ¬y) (¬x y) (¬x ¬y) (¬x ¬y) CNF 2CNF (x y) 2SAT={<φ> | φ ? 2CNF } =(x (y ¬y)) =(x 0) (¬x (y ¬y)) (¬x 0) =x ¬x =0 32 33 3.7 3.8 2. (x y) (x y) (z →(x z) →z P ? ¬y) (z ¬y) (¬w y) (¬w y) (¬x ¬y) L1,L2 (¬x ¬y) (¬w ¬x) P L1 O(nc) A1 L2 O(nd) A2 ¬w A1,A2 A A L1 L2 O(nc+nd)=O(nmax{c,d}) 34 35 5 A1 2SAT A2 B B L1∩L2 O(n(nc+nd))=O(nmax{c,d}+1) A1 C L1 O(nc) C A1 36 37
© Copyright 2024