Rules of the competition :
"Edges have been deleted at random from a citation network. Your mission is to accurately reconstruct the initial network using graph-theoretical, textual, and other information.
In this competition, we define a citation network as a graph where nodes are research papers and there is an edge between two nodes if one of the two papers cite the other."
Best score obtained : 0.97506
Description of the files :
data : repository with the material provided for the competition data_npy : repository that contains the bumpy arrays used for the best submission predictions : cdv files of the two best predictions we made creation_features.py : file which creates and saves the features and label in numpy arrays prediction.py : file which computes the prediction from the numpy arrays saved by creation_features.py
If you want to test our codes you can either :
1 - First run "creation_features.py" to create the training and testing features from the original data in "data". Then, run "prediction.py" which makes a prediction with the features saved by creation_features.npy in data_npy. "prediction.py" writes the file predictions.csv)
2 - You can run directly "prediction.py" with the numpy arrays we provided in data_npy