CD-BPR

A simple yet effective interpretable Bayesian Personalized Ranking for the Assessment of Psychiatric Disorders, 2024

The current repository contains all the code and data necessary to reproduce the paper results.

Installation

Easily clone the repository with git and set up the environment (data and libraries) with make command :

git clone https://gitlab.liris.cnrs.fr/abatel/cd-bpr.git
cd path/to/code
make

Prerequisite : conda library
Attention: It is recommended to uninstall the pytorch library provided with the conda environment to download yourself the version most adapted to your machine from the pytorch website.

Reproductibility

All the commands needed to reproduce the paper results are written in the Experiments.ipynb jupyter notebook at the root of the directory.

Authors

C. Robardet, A. Batel, M. Plantevit, I. Benouaret

LIRIS laboratory, UMR5205, Univ Lyon, CNRS, UBL,LIRIS EPITA Research Laboratory (LRE)