home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!munnari.oz.au!bunyip.cc.uq.oz.au!uqcspe!cs.uq.oz.au!bakker
- From: bakker@cs.uq.oz.au (Paultje Bakker)
- Newsgroups: comp.ai.neural-nets
- Subject: Does anyone have bp on a Super Computer?
- Message-ID: <10107@uqcspe.cs.uq.oz.au>
- Date: 3 Sep 92 06:37:57 GMT
- Sender: news@cs.uq.oz.au
- Reply-To: bakker@cs.uq.oz.au
- Organization: Depts. of Computer Science & Psychology, The University of Queensland, Brisbane, Australia
- Lines: 24
-
- Netters,
- I would like to get in contact with any researcher who has
- been able to successfully implement (or is simply using)
- Rumelhart & McClelland's `bp' simulator on the Connection Machine
- Supercomputer, or the MasPar.
-
- I have obtained an adaptation of the bp program by Sebastian Thrun
- for the Connection Machine (written in 1990) but have been unable to get
- it to work properly; the code appears to have a few bugs, only some
- of which I have been able to find. It does run, but the output
- from parallel computation is different than that of the serial
- computation.
-
- *Please* email me if you can help. Email me also if you are interested
- in the implementation I have already.
-
- thanks,
- Paul Bakker
- University of Queensland
- --
- (paul) bakker@cs.uq.oz.au > "Although the Australian accent does
- computer science/ psychology > not lend itself as a pointer to
- university of qld 4072 > intellect, ... "
- terra australis > The (British) Sunday Telegraph, 1992
-