Newsgroups: comp.constraints
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!cs.utexas.edu!news.sprintlink.net!uunet!psinntp!nntp.hk.super.net!news.ust.hk!kean
From: kean@cs.ust.hk (Alex Kean)
Subject: Re: Expensive constraint checks
Message-ID: <1995Feb13.030230.2613@uxmail.ust.hk>
Sender: usenet@uxmail.ust.hk (usenet account)
Nntp-Posting-Host: cssu49.cs.ust.hk
Organization: Computer Science, HKUST
References: <D3sqwG.I5L@acsu.buffalo.edu>
Date: Mon, 13 Feb 1995 03:02:30 GMT
Lines: 30

In article <D3sqwG.I5L@acsu.buffalo.edu> rchopra@acsu.buffalo.edu (Rajiv Chopra) writes:
>
>Hi!
>	Has anyone addressed constraint satisfaction problems (assume 
>binary, for now) where the constraints are not enumerated as lists of
>acceptable tuples but implicit relations which might be expensive to verify?
>

I don't think you can find optimal strategy for such problem
unless you know a priori the characteristic of the constraints
in yielding a global solution. As a heuristic, I suppose you
could (1) assign a weight to each constraint, and (2) try the 
"cheapest" constraint first and hope that it will reduce the
domain for the "expensive" constraint to verify.


I personally believe this problem is very important because
the "preference ordering" of constraints, whether is for 
optimization or being imposed by the problem, can vastly
alter the strategy for constraint satisfaction. 

Alex Kean
-- 
-------------------------------------------------------------------------
Alex Kean ( ²Ż)
Department of Computer Science                      Email: kean@cs.ust.hk
The Hong Kong University of Science & Technology    Tel: (852) 235 869 92
Clear Water Bay, Kowloon, HONG KONG                 Fax: (852) 235 814 77
-------------------------------------------------------------------------

