A little albeit perhaps rough error study, while we are on the topic:
For completeness, i also ran comparison of sequential svd results
(ssvd.R prototype) on Ahmed's data (this uses normal sampler) and the
MR version.
I actually took time and added option of 0-mean uniform to sequential
R ssvd
If I have a books example , user - book - rating.
And I increase the ratings of books based on their selling , would
that be model-based , since it changed the users ratings ?
Thanks
On Mon, Sep 3, 2012 at 10:25 AM, Sean Owen sro...@gmail.com wrote:
There are a lot of terms for these. Yes they
A quick t-test on these differences gives the same results no
significant difference.
On Mon, Sep 3, 2012 at 11:34 PM, Dmitriy Lyubimov dlie...@gmail.com wrote:
Then i subtracted error means between two methods (+ sign means
smaller error for MR version, -sign means smaller error for R
Hi,
I'm wondering if any has any rules of thumb around model size and memory usage
for SGD? I'm doing some testing of it myself, but thought I would ask to see
how it compares.
Thanks,
Grant