370 likes | 378 Views
Explore variable elimination techniques for optimization and learning in general networks. Understand coherence, difficulty, intelligence, grade, SAT, letter, job, and happiness variables. Discover the process of eliminating variables efficiently, along with applying different ordering strategies and induced graph interpretation. Learn about factorization, heuristic methods, search algorithms, and key theorems for determining optimal orderings.
E N D
CS b553: Algorithms for Optimization and Learning Variable Elimination
Last Time • Variable elimination on polytrees • Top down inference • Linear in size of network • Variable elimination in general • No guarantees… • NP hard in worst case… but when?
Variable Elimination in General Networks Coherence Difficulty Intelligence Grade SAT Letter Job Happy
Variable Elimination in General Networks Coherence Difficulty Intelligence Grade SAT Letter Job Happy
Joint distribution • P(X) = P(C)P(D|C)P(I)P(G|I,D)P(S|I)P(L|G) P(J|L,S)P(H|G,J) • Apply elimination ordering C,D,I,H,G,S,L
Going through VE • P(X) = P(C)P(D|C)P(I)P(G|I,D)P(S|I)P(L|G) P(J|L,S)P(H|G,J) • Apply elimination ordering C,D,I,H,G,S,L • 1(D)=SCP(C)P(D|C)
Going through VE • SCP(X) = 1(D)P(I)P(G|I,D)P(S|I)P(L|G) P(J|L,S)P(H|G,J) • Apply elimination ordering C,D,I,H,G,S,L • 1(D)=SCP(C)P(D|C)
Going through VE • SCP(X) = 1(D)P(I)P(G|I,D)P(S|I)P(L|G) P(J|L,S)P(H|G,J) • Apply elimination ordering C,D,I,H,G,S,L • 2(G,I)=SD1(D)P(G|I,D)
Going through VE • SC,DP(X) = 2(G,I)P(I)P(S|I)P(L|G) P(J|L,S)P(H|G,J) • Apply elimination ordering C,D,I,H,G,S,L • 2(G,I)=SD1(D)P(G|I,D)
Going through VE • SC,DP(X) = 2(G,I)P(I)P(S|I)P(L|G) P(J|L,S)P(H|G,J) • Apply elimination ordering C,D,I,H,G,S,L • 3(G,S)=SI2(G,I)P(I)P(S|I)
Going through VE • SC,D,IP(X) = 3(G,S)P(L|G)P(J|L,S)P(H|G,J) • Apply elimination ordering C,D,I,H,G,S,L • 3(G,S)=SI2(G,I)P(I)P(S|I)
Going through VE • SC,D,IP(X) = 3(G,S)P(L|G)P(J|L,S)P(H|G,J) • Apply elimination ordering C,D,I,H,G,S,L • 4(G,J)=SHP(H|G,J) What values does this factor store?
Going through VE • SC,D,I,HP(X) = 3(G,S)P(L|G)P(J|L,S)4(G,J) • Apply elimination ordering C,D,I,H,G,S,L • 4(G,J)=SHP(H|G,J)
Going through VE • SC,D,I,HP(X) = 3(G,S)P(L|G)P(J|L,S)4(G,J) • Apply elimination ordering C,D,I,H,G,S,L • 5(S,L,J)=SG3(G,S)P(L|G)4(G,J)
Going through VE • SC,D,I,H,GP(X) = 5(S,L,J)P(J|L,S) • Apply elimination ordering C,D,I,H,G,S,L • 5(S,L,J)=SG3(G,S)P(L|G)4(G,J)
Going through VE • SC,D,I,H,GP(X) = 5(S,L,J)P(J|L,S) • Apply elimination ordering C,D,I,H,G,S,L • 6(L,J)=SS 5(S,L,J)P(J|L,S)
Going through VE • SC,D,I,H,G,SP(X) = 6(L,J) • Apply elimination ordering C,D,I,H,G,S,L • 6(L,J)=SS 5(S,L,J)
Going through VE • SC,D,I,H,G,SP(X) = 6(L,J) • Apply elimination ordering C,D,I,H,G,S,L • 7(J)=SL 6(S,L)
Going through VE • SC,D,I,H,G,S,LP(X) = 7(J) • Apply elimination ordering C,D,I,H,G,S,L • 7(J)=SL 6(L,J)
Comparing Orderings • Consider G,I,S,L,H,C,D
Understanding VE: From BNs to Undirected Graphs • Consider each factor as a variable i • Draw an edge between any variables appearing in the same factor
Building the Undirected Graph P(C) Coherence P(I) P(D|C) Difficulty Intelligence P(S|I) P(G|I,D) Grade SAT P(L|G) Letter P(J|S,L) Job Happy P(H|G,J)
Building the Undirected Graph P(C) Coherence P(I) P(D|C) Difficulty Intelligence P(S|I) P(G|I,D) Grade SAT P(L|G) Letter P(J|S,L) Job Happy P(H|G,J)
Building the Undirected Graph Coherence Difficulty Intelligence Grade SAT Letter Job Happy
Variable Elimination Coherence Difficulty Intelligence Grade SAT Letter Job Happy
Variable Elimination Difficulty Intelligence Grade SAT Letter Job Happy
Variable Elimination Difficulty Intelligence Grade SAT Letter Job Happy
Variable Elimination Intelligence Grade SAT Letter Job Happy
Variable Elimination Intelligence Grade SAT Letter Job Happy
Variable Elimination New fill edge Grade SAT Letter Job Happy
Variable Elimination Grade SAT Letter Job Happy
Variable Elimination Grade SAT Letter Job
Variable Elimination Grade SAT Letter Job
Variable Elimination SAT Letter Job
Induced Graph from a VE ordering Coherence Difficulty Intelligence Grade SAT Letter Job Happy
Induced Graph from a VE ordering Coherence Difficulty Intelligence Grade SAT • Theorem: • The scope of every intermediate factor in VE is a clique in the induced graph • Every maximal clique in the induced graph is the scope of an intermediate factor Letter Job Happy
Determining Optimal orderings • Again, NP hard! • Good heuristics in practice: • Min-neighbors, min-fill, etc • Search among elimination orderings while counting size of introduced factors • Greedy search often works well