On Mon, Feb 05, 2018 at 11:58:31AM +0000, Wenhao Huang wrote: > Hi, > > My name is Wenhao and I am a senior computer science student at Peking > University (Beijing, China). I plan to apply for GSoC and contribute to > mlpack library so I guess it's better to start early. > > I took several courses on Machine Learning and I have research experience > concerning application of machine learning on knowledge base. And I > interned at an AI startup focused on CV for 4 months. C/C++ and python are > the languages I use most. > > Recently I am interested in recommender systems and I took a course on > collaborative filtering. Thus I want to work on the CF project for GSoC > 2018. So far I have installed mlpack and successfully run a few programs. > Can someone give me some pointers on what to do next to get started on the > CF idea? Thanks!
Hi Wenhao, I think the best thing to do would be to familiarize yourself with the mlpack code. I might suggest building a CF model in C++ that uses different factorizers, or, perhaps, tuning the parameters of a CF<> model you have built to give maximum RMSE on something like the GroupLens dataset. Once you are familiar with the code, you might look at discussions from years past on this project. Here's an example search that returns some results: https://www.google.com/search?ei=8rx4WqrsI6SpjwSMp56IAg&q=site%3Aknife.lugatgt.org+collaborative+filtering&oq=site%3Aknife.lugatgt.org+collaborative+filtering&gs_l=psy-ab.3...3125.4482.0.4529.17.14.0.0.0.0.128.1043.9j4.13.0....0...1c.1.64.psy-ab..11.0.0....0.csynPi-oRIc (and many of those results will themselves be linked to other posts) You can also browse the archives manually: http://knife.lugatgt.org/pipermail/mlpack/ I hope this is helpful! Thanks, Ryan -- Ryan Curtin | "Avoid the planet Earth at all costs." [email protected] | - The President _______________________________________________ mlpack mailing list [email protected] http://knife.lugatgt.org/cgi-bin/mailman/listinfo/mlpack
