Much like the presentation by Lei Tang (Wallmart Labs) on Adaptive User Segmentation for Recommendation at last year's GraphLab 2013 (see Slides (pdf) here and video here). Xavier Amatriain, of Netflix, made a presentation of what we should be expecting in terms of recommendation. The idea here is that most of this work cannot be static otherwise your customers just won't be responsive to it. Here are his slides and the attendant videos from the Machine Learning Summer School organized in Pittsburgh 2014 by Alex Smola. I note the focus put on matrix and tensor factorizations and the persistent reference to blog posts. It's a new world...more on that later.
Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.