Low RAMP is a Low rank matrix factorization solver based on Belief-Propagation written for Matlab and Julia
It is based on the equations in http://arxiv.org/abs/1503.00338 (Presented in ISIT 2015) and http://arxiv.org/abs/1507.03857 (Presented in Allerton 2015). This work follows and extend a series of older works on approximate message passing for low rank factorization, such as http://arxiv.org/abs/1202.2759,,http://arxiv.org/abs/1402.2238 and NIPS 2014.
Comments and remarks regarding bugs or functionalities are more than welcome, participation is also appreciated.
Having trouble? Send a mail to florent krzakala at firstname.lastname@example.org