Pages: [1]
  Print  
Author Topic: Spatial Clustering  (Read 1279 times)
maccten
Newbie
*
Posts: 28


« on: May 03, 2013, 09:35:23 AM »

Hi All,

This is more of a theoretical question more than anything else
I was wondering if it is possible to use a normal K-NN modeller within rapid miner to cluster data that has latitude longitude associated with it
For example, I have a data set that looks like below where X and Y are latitude and longitude co-ordinates. Where the same ID can have multiple events in multiple locations
Is there a way to make rapid miner see the lat,long information as a  spatial variable. If it looks at it as a regular variable it will give far too much weight to the co-ordinates and a degree of latitude isn't the same as a degree of longitude

    Event 1, Event2, Event 3, Lat Long
ID     0            1          0           X     Y

Thanks very much for you help
Logged
Nils
Administrator
Sr. Member
*****
Posts: 412


« Reply #1 on: May 04, 2013, 04:06:07 PM »

Hi,

before clustering your data you should always normalize it, e.g. with a Z-Transformation.
This will transform it so every attribute has a mean of 0 and a variance of 1.

Best,
Nils
Logged

Please add [SOLVED] to the topic title when your problem has been solved! (do so by editing the first post in the thread and modifying the title)
Pages: [1]
  Print  
 
Jump to: