About time-svd++ implementation

[bang le hai]

Hi, I'm Bang,

I have a question about the implementation of time-svd++ algorithm.

In the term which represents the perspective of implicit feedback in formula (14) - paper "Koren - Collaborative filtering with temporal dynamics",

as my understand, y_j should be a D-dimensional vector (D is the number of latent features) but in the time-svd++ implementation of graphchi, y_j is scalar.

Is this implementation of time-svd++ the same with the original implementation which can get the results reported in the paper?

Thank you,

Bang

[Danny Bickson]

Hi Bang, 

I think you are right, our implementation is a slight variation of the paper since we have a third term in the prediction which is the https://github.com/GraphChi/graphchi-cpp/blob/master/toolkits/collaborative_filtering/timesvdpp.cpp#L168

I guess you can comment this line if you like to return to the original paper formulation.

Best, 

Danny Bickson

Co-Founder
GraphLab Inc.

--
You received this message because you are subscribed to the Google Groups "graphchi-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to graphchi-discu...@googlegroups.com.
To post to this group, send email to graphchi...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/graphchi-discuss/bcb9edd9-1045-4829-bf0a-80366b3fda04%40googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.