/datarobot-mlbench

Evaluation of the DataRobot platform on the mlbench benchmark [H. Zhang et al., 2017]

Primary LanguagePython

datarobot-mlbench

Runs the Kaggle datasets from the mlbench benchmark [H. Zhang et al., 2017] in DataRobot cloud, using the DataRobot API client. Creates multiple submission files for each dataset.

How to Use

  1. Download the mlbench dataset and extract it into this directory.
  2. Configure your access credentials for the DataRobot API.
  3. Run mlbench_preprocess.py.
  4. Run mlbench_run.py.

The script will run all datasets in DataRobot and create 4 submission files for each dataset. These submission files can be uploaded to Kaggle for scoring.