hiroyuki-kasai/Low-Rank-Tensor-Completion

A Dual Framework for Trace Norm Regularized Low-rank Tensor Completion

This package contains a MATLAB implementation of the algorithm presented in the report "A Dual Framework for Trace Norm Regularized Low-rank Tensor Completion" by Madhav Nimishakavi, Pratik Jawanpuria, and Bamdev Mishra; NeurIPS, 2018.

For queries and other related issues, please contact Pratik Jawanpuria (pratik.iitb@gmail.com) or Bamdev Mishra (bamdevm@gmail.com).

Installation:

1. Run run_me_first.m (needed every session).
2. Run compile_mex_c_files.m (needed to be done only once).
3. Run test_Ribeira.m. There are plots at the end.

The proposed files are under the folder 'proposed'.

Disclaimer:

• All files are written by the authors except

  • certain initial mex files were obtained from the works of Nicolas Boumal (nicolasboumal@gmail.com) and (michael.steinlechner@epfl.ch).

• Manopt is downloaded from http://manopt.org.

References:

Please cite the following work if you find the resources in this repository useful.

Madhav Nimishakavi, Pratik Jawanpuria, and Bamdev Mishra. A Dual Framework for Trace Norm Regularized Low-rank Tensor Completion. In Conference on Neural Information Processing Systems (NeurIPS), 2018.

@inproceedings{nimishakavi18,
  title={A Dual Framework for Trace Norm Regularized Low-rank Tensor Completion},
  author={Nimishakavi, Madhav and Jawanpuria, Pratik and Mishra, Bamdev},
  booktitle={Conference on Neural Information Processing Systems (NeurIPS)},
  year={2018}
}

License:

(c) 2017-2018 Madhav Nimishakavi (madhav@iisc.ac.in), Pratik Jawanpuria (pratik.iitb@gmail.com), and Bamdev Mishra (bamdevm@gmail.com).