views:

32

answers:

1

Hi I'm new to Weka, using it to analyse the user attributes based on user ID. the raw data may looks like this,

[userid->game coin]

10001-> 100

10002-> 501

...

i am trying to do a K-Mean Clustering on [game coin] and sort the data into some groups and, is it possible to save the sorted [userid] results, just as some non-overlapped collections of [userid] in Weka? Thanks for your information.

A: 

...Aha, maybe you don't quite understand what I'm talking about, but never mind, I just want to save the clustering result as some collections of elements. Finally I handcrafted my own clustering based on KMeans++ algorithm, in Java, supporting multi-core processing and able to export elements (with id) to CSV table in Mysql... luckily, it works fine...

Xianli