Re: DM: SMP and/or MPP Data Mining Software

[Charlie Berger]

You might like to check out our web page for Darwin.  It runs under 
SMP and
MPP architectures and supports CARTŪ , neural nets, and k-nearest 
neighbor.
It is a scalable, multi-algorithm, client/server data mining software
application designed for large, industrial data mining problems.  

You can also export your models in C, C++ or Java to deploy the 
models into
legacy systems or web-based applications.  Darwin Release 2.2 runs in 
a
client/server architecture with a new WindowsNT/95 client and  UNIX 
(Sun
Solaris or IBM AIX) serial or parallel servers.

Charles Berger
Director of Product Management
Thinking Machines Corporation
617-276-0400 X5020
http://www.think.com




At 04:12 PM 9/19/97 +0100, Chris Nelson wrote:
>You might like to check out our web page for details of 
>Decisionhouse, our
>scalable customer modelling and data analysis software. It supports 
>data
>visualisation, data pre-processing, decision trees, and scorecards 
>amongst
>other things. It can analyse millions of records and runs on parallel
>hardware under UNIX.
>
>Chris Nelson,
>Quadstone Ltd,                                    Tel: +44 131 220 
>4491
>16 Chester Street,                                FAX: +44 131 220 
>4492
>Edinburgh EH3 7RA                           http://www.quadstone.com
>
>EdGercas wrote:
>
>> Are you aware of products that take advantage of SMP or MPP
>> architectures, especially those that offer some advantages in
>> accessing/preparing data from rdb's (Teradata in particular).
>>
>> I am aware of the Neovista Decision Series, the SAS family of 
>products,
>> and ANGOSS's Knowledge Seeker IV.  NCR has some mining tools.  Any
>> others that you are of interest?
>>
>> Also, any white papers that address the leveraging of these
>> architectures would be very much appreciated.
>>
>> Thanks.
>
>
>
= = = = = = = = = = = = = = = = = 
Charlie Berger
Thinking Machines Corporation
14 Crosby Drive
Bedford, MA  01730
(617) 276-0400
= = = = = = = = = = = = = = = = =