iarai / science4cast Goto Github PK

I was checking the calculate_ROC function in the tutorial against sklearn.metrics.roc_auc_score and they do not seem to produce the same value. Could you please take a look at this?

Thank you!

Thank you for organizing this competition.
I am very much looking forward to the youtube video of the workshop talks.

About the data from 2017-2020, do you have plan to release these data as well?

First, get the data from the IARAI cloud. It contains three files:

• TrainSet2014_3.pkl: (33 MB) Semantic network until 2014, and list of unconnected vertex-pairs that can be predicted for training the neural network.
• TrainSet2014_3_solution.pkl: (0.5 MB) Solution of whether the unconnected vertex pair from 2014 has been connected by 2017.
• CompetitionSet2017_3.pkl: (95 MB) Semantic network until 2014, and list of unconnected vertex-pairs that should be predicted for the evaluation.

That last line should be the Semantic network until 2017, I think.

From what I understood, vertices mentioned in unconnected_vertex_pairs, should also be present in the vertices mentioned in full_dynamic_graph_sparse. But, after running following snippet,

u_v = set(unconnected_vertex_pairs[:, 0]) | set(unconnected_vertex_pairs[:, 1])
c_v = set(full_dynamic_graph_sparse[:, 0]) | set(full_dynamic_graph_sparse[:, 1])

the output of len(u_v - c_v) is 13152. That means, 13752 vertices in the unconnected list do not appear in the full dynamic graph. I may understood it wrong, so please clarify.