15.08.2013 Views

General Computer Science 320201 GenCS I & II Lecture ... - Kwarc

General Computer Science 320201 GenCS I & II Lecture ... - Kwarc

General Computer Science 320201 GenCS I & II Lecture ... - Kwarc

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Example 613 If the rules of the 8-puzzle are relaxed so that a tile can move anywhere,<br />

then we get heuristic h1.<br />

Example 614 If the rules are relaxed so that a tile can move to any adjacent square, then<br />

we get heuristic h2.<br />

Key point: The optimal solution cost of a relaxed problem is not greater than the optimal<br />

solution cost of the real problem<br />

c○: Michael Kohlhase 491<br />

Relaxation means to remove some of the constraints or requirements of the original problem, so<br />

that a solution becomes easy to find. Then the cost of this easy solution can be used as an<br />

optimistic approximation of the problem.<br />

4.1.5 Local Search<br />

Local Search Problems<br />

Idea: Sometimes the path to the solution is irrelevant<br />

Example 615 (8 Queens Problem)<br />

Place 8 queens on a chess board, so that<br />

no two queens threaten each other.<br />

This problem has various solutions, e.g.<br />

the one on the right<br />

Definition 616 A local search algorithm<br />

is a search algorithm that operates<br />

on a single state, the current<br />

state (rather than multiple paths).<br />

(advantage: constant space)<br />

Typically local search algorithms only move to successors of the current state, and do not<br />

retain search paths.<br />

Applications include: integrated circuit design, factory-floor layout, job-shop scheduling, portfolio<br />

management, fleet deployment,. . .<br />

c○: Michael Kohlhase 492<br />

Local Search: Iterative improvement algorithms<br />

Definition 617 (Traveling Salesman Problem) Find shortest trip through set of cities<br />

such that each city is visited exactly once.<br />

Idea: Start with any complete tour, perform pairwise exchanges<br />

259

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!