home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.lang.prolog
- Path: sparky!uunet!math.fu-berlin.de!mailgzrz.TU-Berlin.DE!cs.tu-berlin.de!hoppet
- From: hoppet@opal.cs.tu-berlin.de (Thomas Hoppe)
- Subject: Re: Machine Learning Program Library
- Message-ID: <1993Jan8.121958.7970@cs.tu-berlin.de>
- Keywords: Machine Learning
- Sender: news@cs.tu-berlin.de
- Organization: Techn. University of Berlin, Germany
- References: <emde.726485815@hercules> <1993Jan8.113247.18728@dcs.qmw.ac.uk>
- Distribution: comp.lang.prolog
- Date: Fri, 8 Jan 1993 12:19:58 GMT
- Lines: 29
-
- mmh@dcs.qmw.ac.uk (Matthew Huntbach) writes:
-
- >You ought to have Shapiro's model inference system. I am not
- >sure if the code is available anywhere, but the text is given
- >as an appendix to his thesis "Algorithmic Program Debugging"
- >which is published in MIT's Distinguished Dissertation series.
- >I wrote an improved version of it, but I used the Poplog
- >system, mixing Pop-11 with the Prolog, so it wouldn't be
- >translatable.
- >
- >Matthew Huntbach
-
- Yes, that's right. Actually I have re-implemented it already some
- years ago and it is still on my stack of programs "to be included in
- the lib". Since the behavior of the re-implementation differs from the
- behavior published in Shapiro's book and since I have compared the
- code over and over again, it seems that the behavior difference is a
- result of different Prolog dialects. Thus and because of a lack of
- time this program isn't yet in the Lib. Anyway, if I find some time in
- the future it will surely be included.
-
- If somebody is interested in it, s/he can mail me directly to get it.
-
- Thomas Hoppe
- --
- Thomas Hoppe Projekt KIT-BACK email: hoppet@cs.tu-berlin.de
- Technische Universitaet Berlin phone: 030-314-25494
- Franklinstr. 28/29 fax : 030-314-24929
- 1000 Berlin 10 Germany priv.: 030-3236564
-