Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
a = 0 a = 1
b = 0 b = 1 b = 0 b = 1
c = 0 c = 1 c = 0 c = 1 c = 0 c = 1 c = 0 c = 1
In pure CSP there is no optimization But this does not mean you
function! (think of the Crystal Maze)
© J. Christopher Beck 2015 can’t use CP for optimization!5
Standard CP Search
n Branch-and-Infer
n at each node, run constraint propagation
n For Constraint Optimization Problems:
n Branch-Infer-&-Bound
n Partition/Inference/Relax (see Lecture 10)
n Often bound calculation is “hidden” in
constraint propagation
A1 15
A2 20
35 50 100 120
MinSlack Heuristic
n Find operation pair with smallest slack
n Post the opposite sequence
n Example
n slack(A → B) = 50, slack(B → A) = 25
n slack(A → C) = 150, slack(C → A) = 5
n slack(B →
C) = 15, slack(C → B) = 50
A1 → A2 A2 → A1
A4 →
A1
…