Hi, I would also like to give a hand in the development of Spark, and more especially on the Mllib subproject. There are plenty of things we could port from other similar projects (mahout, graphlab, scikit-learn), but I'm pretty sure someone has already thought about all of this.
As a starting point I was thinking about implementing some evaluation methods for recommender systems (RMSE,NDCG or more simply hit rate@k and precision@k). I suspect there is someone I should contact before writing my stuff and submitting a pull request ? Guillaume Pitel - eXenSa Josh Rosen <rosenvi...@gmail.com> a écrit : >Check out the replies in this thread: >https://mail-archives.apache.org/mod_mbox/spark-dev/201310.mbox/%3ccaasvfpnhwnx0shhzklcz9_kfsw_6hxtwrwoxajsojcdgkiq...@mail.gmail.com%3E > > >On Wed, Oct 23, 2013 at 1:21 AM, 周杰 <zhoujie...@126.com> wrote: > >> Hi, I want to know how to contribute to spark
