MIT15_053S13_iprefguide

In each case we need bounds on how much the

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: a constraint is satisfied, and 0 otherwise. In each case, we need bounds on how much the constraint could be violated in a solution that satisfies every other constraint of the problem. Example 1. ⎧1 ⎪ w= ⎨ ⎪0 ⎩ if x ≥ 1 if x = 0. In this example, x is an integer variable. And suppose that we know x is bounded above by 100. (We may know the bound on x because it is one of the constraints of the problem. We may also know the bound of 100 on x because it is implied by one or more of the other constraints. For example, suppose that one of the constraints was "3x + 4y + w : 300." We could infer from this constraint that 3x : 300 and thus x : 100. Equivalent constraint: w : x : 100w. w E {0,1}. In any feasible solution, the definition of w is correct. If x � 1, then the first constraint is satisfied whether w = 0 or w = 1, and the second constraint forces w to be 1. If x = 0, then the first constraint forces w to be 0, and the second constraint is satisfied. ⎧1 ⎪ if x ≥ 7 Example 2. w = ⎨ otherwise. ⎪ 0 ⎩ Agai...
View Full Document

Ask a homework question - tutors are online