Hi,

 

I am at a beginner level in using Mahout and m planning to build a random forest model on Customer data  to classify churners and non-churners.

Any help on as to how can we go about it in Mahout?

Does any parallelized algorithm for Random forest is available?

Will Naïve Bayes/Complementary Naïve Bayes work in this situation?

 

 

Thanks,

Akshay Nowal

 

Mahout Random Forest does not seems scale out well. See a recently PAKDD paper on some new implementation based on hadoop:
" Scalable Random Forests for Massive Data.pdf "
 
--Tao

-----Original Message-----
From: Nowal, Akshay [mailto:[EMAIL PROTECTED]]
Sent: 2012年4月11日 14:12
To: [EMAIL PROTECTED]
Subject: How to use Random forest in Mahout

Hi,

 

I am at a beginner level in using Mahout and m planning to build a random forest model on Customer data  to classify churners and non-churners.

Any help on as to how can we go about it in Mahout?

Does any parallelized algorithm for Random forest is available?

Will Naïve Bayes/Complementary Naïve Bayes work in this situation?

 

 

Thanks,

Akshay Nowal