Aymane BERRADI - Taoufik AGHRIS - Badr LAAJAJ
It contains the different python files that include useful functions and classes.
It contains a text file of drive link to have all necessary data
Text_Preprocessing_and_Embbeding.ipynb
: it contains all preprocessing steps for text data, and embedding using Doc2Vec and S-BERT.Node_embedding.ipynb
: it contains the embedding of the nodes using deepwalk for weighted and unweighted versions of the graph.Features_engineering.ipynb
: it contains the structural features for graph data, and constructed variables used in theAltegrad_Models.ipynb
.Altegrad_Models.ipynb
: it contains several experiments on both features configuration and predictive models.Kaggle_Submission.ipynb
: it's the notebook that reproduces predictions of our best performing model and lead us to the first place with score of 3.01562 on the private leaderboard https://www.kaggle.com/c/altegrad-2020/leaderboard.