I really need your help..i took a project in my college about data mining,the project is a comparison on a 3 free software for data mining. I choose the first two(WEKA,YALE) but i am searching for the third and i am litle bit confused which to choose..
I found Knime interisting but ii dont know if it is worth to work with it...or if it is on the same area with WEKA and YALE..
Knime is easy on setup???
If you can help me to choose with your personal experience i will be glad...
I am maybe a tiny bit biased here :-)
but... Weka, KNIME, and Yale are probably a nice matchup because they
represent three different paradigms. Weka is more of a state of the art
machine learning toolkit. Yale focusses more on table-base data visualization
and KNIME's stronghold is on ease-of-use and modularity via data pipelining.
Both Yale and KNIME integrate Weka (or parts of it, at least).
Take the above with a grain of salt, however. It's not quite as black&white
as I am painting it.
YALE recently was released w/ new name RapidMiner. It has removed the last vestiges of Weka, I believe. I looked at both of those 2 and found YALE/RapidMiner to have more sophisticated mechanism of authoring complex learning schemes thank Weka. Both Weka and YALE are very general- they can be used on most any (tabular) data. YALE/RapidMiner seems to have very responsive online support.
I am curious to see KNIME compared with these 2. On my first glance it appears to be a specialized tool for bio- or cheminformatics. Could you kindly describe some more general use cases of how KNIME can be used for other types of data?
Thanks!
David
On my first glance it appears to be a specialized tool for bio- or cheminformatics.
far from that - if you look at http://www.knime.org/example_apps.html on this site you'll already see a few other application areas. KNIME is a general purpose data pipelining toolkit, not more, not less. As such it differs from more learning method oriented tools such as Yale and Weka. If you aren't using any, you probably won't appreciate the (dis)advantages of the different approaches...