Netflix_Recommander_System

Implementation of 5 methods of recommander system on Netflix data.

Baseline Estimator ( baseline_estimator.py )
Correlation Based Neighbourhood Model ( correlation_based_neighbourhood_model.py )
Correlation Based Implicit Neighbourhood Model ( correlation_based_implicit_neighbourhood_model.py )
SVD++ ( svd_more_more.py )
Integrated Model ( integrated_model.py )

How to use scripts

The default dataset path is "../Datasets" (relative path from the root of this repository), it can be modified in utils.py.

In this folder it must be a folder named "download" with the content of the archive that can be download here: https://archive.org/download/nf_prize_dataset.tar

The folder is not included in this repository due the size of the files.

Then you can run rating_compiler.py at first to create the different matrices from the files. Then you can run each algorithms.

Name		Name	Last commit message	Last commit date
Latest commit History 25 Commits
.ipynb_checkpoints		.ipynb_checkpoints
AFMatriciel_Project.ipynb		AFMatriciel_Project.ipynb
Factorization_Meets_the_Neighborhood-_a_Multifaceted_Collaborative_Filtering_Model.pdf		Factorization_Meets_the_Neighborhood-_a_Multifaceted_Collaborative_Filtering_Model.pdf
MLDS_AFMatriciel_PONT_RODRIGUES.tar.gz		MLDS_AFMatriciel_PONT_RODRIGUES.tar.gz
Projet_Factorisation_et_Apprentissage.ipynb		Projet_Factorisation_et_Apprentissage.ipynb
README.md		README.md
baseline_estimator.py		baseline_estimator.py
correlation_based_implicit_neighbourhood_model.py		correlation_based_implicit_neighbourhood_model.py
correlation_based_neighbourhood_model.py		correlation_based_neighbourhood_model.py
integrated_model.py		integrated_model.py
rating_compiler.py		rating_compiler.py
svd_more_more.py		svd_more_more.py
utils.py		utils.py