Pages: [1]
  Print  
Author Topic: RapidMiner 5 Beta released  (Read 3087 times)
Ingo Mierswa
Administrator
Hero Member
*****
Posts: 1179



WWW
« on: October 06, 2009, 12:02:02 PM »

Hello.

Finally, Rapid-I presents the new version of RapidMiner. Among the new features of RapidMiner 5 are a new user interface, the new repositories, a completely new way of meta data transformation, and the quick fixes, which simplifies the design of complex analysis processes a lot.


Please find a first overview in our video: „RapidMiner 5 in 3 Minutes“:

http://rapid-i.com/videos/rm_5_demo_EN/rm_5_demo_EN.html




More information about the RapidMiner 5 Beta release can be found at

http://rapid-i.com/content/view/172/1/


Please let us know what you think in our new RapidMiner 5 Beta board here in this forum.

Cheers,
Ingo
Logged

Did you try our new Marketplace? Upload or download new Extensions, add comments, and organize your operators. Have a look at  http://marketplace.rapid-i.com
crappy_viking
Newbie
*
Posts: 18


« Reply #1 on: October 08, 2009, 01:22:56 PM »

Hi,

I saw that there are graphs in rapidMiner 5.0. Is that like Cytoscape ? (www.cytoscape.org)
I think Cytoscape is great with a plugin called "network inference" about genes and proteines. If features and attributes in RM are thought of as genes, inferring a gene network would be like inferring a partial correlation network between features : In the "filter" approach, exploring links between attributes would be awesome !!
Here are a few references :
- www.cytoscape.org
- "GeneNet" in R-CRAN for "R" language
Logged
Sebastian Land
Administrator
Hero Member
*****
Posts: 2416


« Reply #2 on: October 09, 2009, 08:08:29 AM »

Hi,
at least some of the screenshots make me think, that both programs are using the same graph library. But RapidMiner focuses much stronger on general data mining and is not specialized on gene networks. I think you might get similar or even better results with the data mining functionality of RapidMiner, but it will show the results in a less appealing way.
Of course one could think of a plugin for rapid miner, giving more specialized operators for bio informatics and visualizations, but this would be a greater project and we would need a partner from the biological side, because we don't have a clue about the demands in this field.

Greetings,
  Sebastian
Logged

Hope to see you at RapidMiner Community Meeting and Conference (RCOMM 2011) in Dublin from June 7-10, 2011.
The Call for Paper is online now!
More information at http://www.rcomm2011.org
crappy_viking
Newbie
*
Posts: 18


« Reply #3 on: October 10, 2009, 11:28:38 AM »

Hi Sebastian Land, and thank you for replying.

It was not exactly what I wanted to say. The point was to reuse the mathematical aspect of an existing approach (gene analysis) to fulfill data mining tasks, to say, feature selection. You really can compute a graph between features, using one of these techniques, the choice will be the computational load :

Imagine now that you have an example set with 100 attributes :
row attr1 attr2 ...attr100
1      0.5   3    ....2
2      -0.7  4     ...1
3      0      4    ... 3
....

With such a technique, you should obtain a graph between features, with the following adjacency matrix :
         attr1 attr2 attr3 ...
attr1    0      1     0     ...
attr2    1      0     0
attr3    1      1     0     ...
etc...

Thus, manipulating connected components (and others), you could detect which attributes are relevant. There "filter approach" would be powerful, both for data quality/profiling and for preprocessing.

About Cytoscape, actually two plugins would be needed, not exactly the whole software :
- Network Inference : http://www.baderlab.org/Software/ExpressionCorrelation
- Network clusters : http://baderlab.org/Software/MCODE
Logged
Sebastian Land
Administrator
Hero Member
*****
Posts: 2416


« Reply #4 on: October 12, 2009, 08:44:15 AM »

Hi again.
Then you are suggesting that we implement this feature selection methods? I would have to take a closer look at each of the algorithms for calculating the time we would need, but unfortunately I can already say, that we will not find the time this year. (Paid) customer projects always have priority and there are many things needed to accomplish for the final version of RapidMiner 5. But I will keep this in mind, especially because we are thinking of publishing a special feature selection plug-in in near future.

Greetings,
  Sebastian
Logged

Hope to see you at RapidMiner Community Meeting and Conference (RCOMM 2011) in Dublin from June 7-10, 2011.
The Call for Paper is online now!
More information at http://www.rcomm2011.org
Pages: [1]
  Print  
 
Jump to: