Hi Magicalo, You can find a patch for PCA under MAHOUT-512 which is available here https://issues.apache.org/jira/browse/MAHOUT-512.
This implementation scales well with training samples and calculates the covariance matrix in a distributed way. The feature size is not so scalable as the SVD is done sequentially. I believe there is the intent to address large scale feature size by adding PCA to the SSVD as part of MAHOUT-817, perhaps around Q1 2012. On Dec 4, 2011, at 7:58 PM, magicalo <[email protected]> wrote: > Hello, > > Is there an expected release date for the PCA algorithm as part of Mahout? Tx! >
