Benchmarks for Kaggle's Predict Closed Questions on Stack Overflow competition
The benchmarks require several Python packages:
These packages can be installed with easy_install or pip, or Windows users can download compiled versions of these packages
To run the benchmarks, you also need to download the data. The only files necessary for the benchmarks are train-sample.csv and public_leaderboard.csv. Two variables need to be updated in competition_utilities.py as well: data_path should be set to the path to the data, and submissions_path should be set to the location for writing the submission files.