This is the code to reproduce the results of our paper Exploring the Potential of Optimal Active Learning via a Non-myopic Oracle Policy.
Unzip the results.zip folder in the root directory.
To plot the learning curves, use the command python -m evaluation.plot_learning_curve.
To print the win-tie-loss table, use the command python -m evaluation.win_tie_loss_tables.
To print the ranking tables, use the command python -m evaluation.ranking_tables.