Skip to content

Experimental Codes for Collaborative Filtering with Graph Information: Consistency and Scalable Methods.

License

Notifications You must be signed in to change notification settings

rofuyu/exp-grmf-nips15

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Requirement
===========
    - Unix-like Environment
    - Gnu GCC/G++ 4.8 or above 
    - Matlab or Octave 
    - If you are using Mac OS X, please install GCC and Octave via Mac HomeBrew


Install
=======
    - Open a Matlab or Octave interactive terminal
    - type 
        > install

Usage
=====
        >> addpath('trmf-core/matlab') % or addpath('trmf-core/octave');
        >> glr_mf_train                  
        Usage: [W' H' rmse walltime] = glr_mf_train(Y, testY, A, B [,
                'options'])
        Usage: [W' H' rmse walltime] = glr_mf_train(Y, testY, A, B, W', H' [,
                'options'])
		       Y: [I J V] a nnz-by-3 matrix with each row as an entry in the 
			      observed matrix
               A: a symmetric sparse matrix, e.g., Laplacian for rows
               B: a symmetric sparse matrix, e.g., Laplacian for cols
               options:
                -n threads : set the number of threads (default 4)
                -k rank : set the rank (default 10)
                -e epsilon : set stopping criterion epsilon of tron (default 0.1)
                -t max_iter: set the number of iterations (default 10)
                -g max_cg_iter: set the number of iterations used in CG (default 5)
                -q verbose: show information or not (default 1)


FAQ
===
    - If you see something like 
        "Invalid MEX-file ... libstdc++.so.6: version `CXXABI_1.3.8' not found..,"
      the libstdc++ in your system is newer than the version that comes with
      matlab. You can resolve this problem by opening matlab with a LD_PRELOAD
      environment variable such as 
        LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libstdc++.so.6 matlab


Citation
========
Please acknowledge the use of the code with a citation.
@InProceedings{   NR15a,
  title={Collaborative Filtering with Graph Information: Consistency and Scalable Methods},
  author={Rao, Nikhil and Yu, Hsiang-Fu and Ravikumar, Pradeep K. and Dhillon, Inderjit S.},
  booktitle = {Advances in Neural Information Processing Systems 27},
  year={2015}
}


If you have any questions regarding the code, feel free to contact Hsiang-Fu Yu (rofuyu at cs utexas edu). 

About

Experimental Codes for Collaborative Filtering with Graph Information: Consistency and Scalable Methods.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages