Skip to content

abhijithch/RecSys.jl

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

RecSys

Build Status

RecSys.jl is an implementation of the ALS-WR algorithm from "Yunhong Zhou, Dennis Wilkinson, Robert Schreiber and Rong Pan. Large-Scale Parallel Collaborative Filtering for the Netflix Prize. Proceedings of the 4th international conference on Algorithmic Aspects in Information and Management. Shanghai, China pp. 337-348, 2008"

Usage

  • Install: Pkg.clone("https://github.com/abhijithch/RecSys.jl.git")
  • Specify the training dataset in one of several ways:
    • Use delimited (CSV) file with columns: user_id, item_id, ratings. E.g.: trainingset = DlmFile("ratings.csv", ',', true).
    • Use a MAT file, specifying the file and entry name. E.g.: trainingset = MatFile("ratings.mat", "training")
    • Provide an implementation of FileSpec for any other format.
  • Initialize: als = ALSWR(trainingset)
  • Train: train(als, num_iterations, num_factors, lambda)
  • Check model quality:
    • rmse(als) to check against training dataset
    • rmse(als, testdataset) to check against a test dataset
    • and repeat training with different parameters till satisfactory
  • Save model: save(als, filename)
  • Load model: als = load(filename)
  • Get recommendations:
    • recommend(als, user_id) for an existing user
    • recommend(als, user_ratings) for a new/anonymous user

Examples

See examples for more details:

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages