發(fā)信人: yaomc (白頭翁&山東大漢), 信區(qū): DataMining
標(biāo)  題: [合集]時間序列的挖掘
發(fā)信站: 南京大學(xué)小百合站 (Wed Jan 16 17:00:14 2002), 站內(nèi)信件

dewbow (能通六竅) 于Wed Dec 19 20:48:55 2001提到：

我覺得現(xiàn)在挖掘技術(shù)用的最多的是商業(yè)上的應(yīng)用。
但在工業(yè)上，時間序列的挖掘是一個很重要的方面。
但我發(fā)覺一些這里關(guān)于這方面的討論比較少。
我是流程工業(yè)自動化的，我想在時間序列挖掘上
做點事情，請各位高手指點迷經(jīng)啊。


yaomc (白頭翁&山東大漢) 于Wed Dec 19 21:15:30 2001提到：

You can introduce some of the time series mining to us.
Thanks.


fervvac (高遠) 于Wed Dec 19 23:07:53 2001提到：

What I know about sequence data mining:
1. Category:
  +) sequential data mining
  +) episode discovery
  +) periodic pattern mining

2. Major applications includes: web data log mining, stock market mining, 
  DNA sequence mining.

3. Difficult parts:
  +) Pushing constraints into mining, as well as those complex constraints, or
     their combinations.
  +) Mining long patters, "irreglar" patters as well as in-frequent patterns.
  +) Find more applications 

Maybe you can illustrate the problems you encountered in the industry in 
details, in order to have further discussions.