home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!spool.mu.edu!uwm.edu!rutgers!netnews.upenn.edu!gradient.cis.upenn.edu
- From: freeman@gradient.cis.upenn.edu (Jon Freeman)
- Newsgroups: comp.theory
- Subject: Two questions about CSP's
- Message-ID: <87907@netnews.upenn.edu>
- Date: 4 Sep 92 03:04:41 GMT
- Sender: news@netnews.upenn.edu
- Organization: University of Pennsylvania
- Lines: 23
- Nntp-Posting-Host: gradient.cis.upenn.edu
-
- Greetings.
-
- 1. I would like to obtain a list of references that discuss the time
- and space complexity of converting the variables in a constraint
- satisfaction problem (CSP) from non-binary to binary form (i.e.,
- converting a CSP to a SAT problem).
-
- 2. I would also like some pointers to quotes from the literature in
- which researchers express their opinion as to which of these
- representations, if any, is better for a particular CSP technique or
- set of techniques. For example, in their paper "Backtracking with
- Multi-Level Dynamic Search Rearrangement" (Acta Informatica 15:99-113,
- 1981), Purdom, Brown, and Robertson state:
-
- "Transforming a problem into one with binary variables
- gives maximum scope to search rearrangement."
-
- Thanks in advance,
- Jon
- -------------------------------------------------------------------------------
- "If the accelerator (gas pedal) sticks, the | Jon Freeman
- car will often keep going faster and faster." | freeman@gradient.cis.upenn.edu
- --Pennsylvania Driver's Manual, p. 55 | Office: (215) 898-4612
-