6.txt

This complete matlab for neural network

发信人: fervvac (高远), 信区: DataMining
标  题: Re: 请问基于DW的DM与基于DB的DM具体做法有什么不同？
发信站: 南京大学小百合站 (Wed Mar 20 08:46:20 2002), 站内信件

I think one difference might be that once one can leverage the existing data
cube for the mining process. Prof. Jiawei Han has quite a lot of publication
on this aspect. Note that his DM group attempted to integarte DM with DW for
a long time and it seems they are re-focusing on this topic now.

Another thing to mention is that most mining on top of the "relational" data
is actually not so appropriate. They are mostly transactional data. It is 
of course possible to convert transactional dat into relational data, but the
value 0 and * cannot be distinguished (in original dm algorithms).

【 在 helloboy (hello) 的大作中提到: 】
: 好像dW经常要涉及表的连接，在dm处理的时候复杂很多。
: 请问有无这方面的文章？


--
※ 来源:．南京大学小百合站 bbs.nju.edu.cn．[FROM: 饮水思源BBS]