Hadoop extension for RapidMiner

[Preko]

Hello Everyone,

I am happy to announce Radoop, our integration effort between RapidMiner and Hadoop, the most popular distibuted computing environment. If you have ever experienced limitations because of your data not fitting into memory or your processes running too slow on one machine, then Radoop might be the solution for you.
Please sign up at http://radoop.eu and share the word!

I hope to se you at RCOMM 2011 in Dublin, where Radoop will be presented!

Thanks, Zoltan

[earmijo]

Excellent Preko. Thank you very much. Integration with Hadoop is a feature I've been expecting from RapidMIner. I signed up for a test drive but I think you are still in early phases. Do you have any idea when will it be released for the general public?

[wessel]

Nice.

I'm working with Hadoop right now.

Would be nice to see it iterated with Rapid Miner.

But doesn't Hadoop lack a huge amount of algorithms?

https://cwiki.apache.org/confluence/display/MAHOUT/Algorithms

Classification

A general introduction to the most common text classification algorithms can be found at Google Answers: http://answers.google.com/answers/main?cmd=threadview&id=225316 For information on the algorithms implemented in Mahout (or scheduled for implementation) please visit the following pages.

Logistic Regression (SGD)

Bayesian

Support Vector Machines (SVM) (open: MAHOUT-14, MAHOUT-232 and MAHOUT-334)

Perceptron and Winnow (open: MAHOUT-85)

Neural Network (open, but MAHOUT-228 might help)

Random Forests (integrated - MAHOUT-122, MAHOUT-140, MAHOUT-145)

Restricted Boltzmann Machines (open, MAHOUT-375, GSOC2010)

Online Passive Aggressive (awaiting patch commit, MAHOUT-702)

[Preko]

Thank you for the interest!

earmijo: Yes, we are still in heavy development, but the tool is already great for ETL and clustering. I think all beta applicants will get access by the end of July and we might have a full public release in September.

wessel: Of course, Hadoop has less modeling algorithms than RapidMiner as many algorithms are hard (or even impossible) to parallelize effectively. The good news is that if you have huge amount of data, then even simple methods can reach great results. It is usually better to use more data than to use a more sophisticated algorithm.

If you plan to try it out, then apply at http://radoop.eu or send me an email.

[alandhw]

Hi, I having research project by putting rapid miner on cloud computing. I wish to experience the radoop to improve my knowledge on it. where can I get Radoop?

[Preko]

Hi,

You can apply for a beta account at http://radoop.eu
If you have a specific use-case, have a Hadoop cluster, and you are willing to provide feedback while using Radoop, then please describe your use-case in email and we can speed up the process: preko (at) radoop.eu

Thanks

[StaryVena]

Hi,
when will be extension and source codes freely available? After publishing release version?

Cheers
Vaclav

[Preko]

Hi,

The extension should be publicly available in Q1 2012, but we don't know yet about the code.

Best, Zoltan

[alandhw]

Hi, preko. is there the radoop able to deploy on openNebula cloud platform? I have to send email to who in order to get the beta version of radoop for research purpose? Thank you.

[Preko]

Hi alandhw,

As far as I know, openNebula is a general cloud platform, so I am sure that Hadoop can be deployed on it.
Currently, you can only use Radoop if you have a working Hadoop cluster. If you do have one, then please write to preko (at) radoop.eu and describe what you plan to do with Radoop.

Best, Zoltan

[alandhw]

Hi, Preko,

I only allow to apply for beta version through invite more friends to view on it. I could not find the contact email to the host. can u provide me the email address to write in? Thanks

 

[earmijo]

>  then please write to preko (at) radoop.eu and describe what you plan to do with Radoop.

preko@radoop.eu