IMPORTANT: This is the old Data Mining forum.

I keep it online so that you can read the old messages.

Please post your new messages in the

quantitative association rule mining

Posted by:
**
Qi
**

Date: April 08, 2014 04:29PM

Hi,

Common association rule mining algorithm can give us the result a->b from the dataset. What if I want to mining the rules like a>10->b<8 from the dataset?

Please is there any algorithm that solves this problem?

Thanks very much

Qi

Edited 1 time(s). Last edit at 04/08/2014 05:22PM by webmasterphilfv.

Common association rule mining algorithm can give us the result a->b from the dataset. What if I want to mining the rules like a>10->b<8 from the dataset?

Please is there any algorithm that solves this problem?

Thanks very much

Qi

Edited 1 time(s). Last edit at 04/08/2014 05:22PM by webmasterphilfv.

Posted by:
**
webmasterphilfv
**

Date: April 08, 2014 05:21PM

This is called "quantitative association rules" or "numeric association rules". There exists some research papers about this topic proposing some algorithms but I don't know any public source code or implementation.

Posted by:
**
Qi
**

Date: April 09, 2014 07:38AM

Thanks very much. I'm new in this area, at least this is a clue.

Posted by:
**
Qi
**

Date: April 15, 2014 08:21PM

I have a new question in this process. How can we decide the interval of the data. I mean the boundary of intervals. It seems there will be lots of less important rules. Really need to fix this problem...

Thanks,

Qi

Thanks,

Qi

Posted by:
**
webmasterphilfv
**

Date: April 16, 2014 03:59AM

I did not read much about this topic, so I cannot answer this question. But I guess that the algorithms may be able to choose the interval automatically.