LLORMA-tensorflow

• This repository is tensorflow implementation of "Local Low-rank Matrix Approximation".

  • I implemented two version of LLORMA: Parallel LLORMA (ICML'13) and Global LLORMA (JMLR'16).

• I have increased the batch size for performance. If you want to get results as same as the original paper, please set batch size to 1.

• I refer the codes from https://github.com/jnhwkim/PREA/tree/master/src/main/java/prea/recommender/llorma.

• Folder description

  • llorma_p: Parallel LLORMA (ICML'13)
  • llorma_g: Global LLORMA (JMLR'16)

• Dependecy: python3.5 and see requirements.txt

• How to run

python train.py