Music Recommandation System based on Yahoo! music dataset.
--Hierarchy.py generate hierarchy structure for a list of track items.
--Rating.py generate ratings for all the items in the hierarchy structure.
--matrix_factorization.py apply matrix_factorization on datasets to make predictions.
--ADD.py make predictions by simply adding album and artist rating.
--Ensemble.py ensemble all the results obtained from previous steps and make new predictions.
--Report.pdf introduction of methods applied in this project.
Here is a instruction about how to use the programs. All the methods are mentioned in the report.
Methods 1--6
-
Run Hierarchy.py to get the hierarchy structure.
-
Run Rating.py to get ratings for every item of each user.
-
Run Add.py to apply our methods by changing some parameters.
Method 7
Run matrix_factorization.py to perform matrix factorization method by adjusting rank and numiter
Method 8
Run Ensemble.py to perform ensemble method