home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.ai.neural-nets:3198 comp.ai:3103 comp.compression:3021 comp.theory:1766
- Newsgroups: comp.ai.neural-nets,comp.ai,comp.compression,comp.theory
- Path: sparky!uunet!s5!sethb
- From: sethb@fid.morgan.com (Seth Breidbart)
- Subject: Re: measuring complexity (was Re: Kolmogorov Complexity)
- Message-ID: <1992Aug15.025443.18937@fid.morgan.com>
- Organization: Morgan Stanley & Co., New York, NY
- References: <1992Aug10.163725.26986@julian.uwo.ca> <1992Aug10.115352.1@acad3.alaska.edu> <1992Aug14.212005.3438@julian.uwo.ca>
- Date: Sat, 15 Aug 1992 02:54:43 GMT
- Lines: 16
-
- In article <1992Aug14.212005.3438@julian.uwo.ca>
- koops@asterix.gaul.csd.uwo.ca (Luke Koops) writes:
-
- > Also, I beleive that K-complexity is not unbiased, but is
- >biased towards Turing machine representations of concepts. It just
- >provides a really good estimate of complexity. This is merely my
- >opinion, I would appreciate it if someone could confirm it.
-
- K-complexity is defined relative to any universal representation,
- which can be Turing Machines, APL programs, 2-counter machines, ...
- The difference between the complexity of all strings in any two models
- is bounded by a constant. Why do you feel that measuring complexity
- of a string as the length of the shortest C program that produces that
- string to be biased towards Turing Machines?
-
- Seth sethb@fid.morgan.com
-