arXiv (Cornell University) · 2021 · 19 citations · 0 references
Public HealthComputer ScienceCausal InferenceMachine LearningStatistical ModelingCausal Graphical ModelsStatistical Software DevelopmentCausal Machine LearningCausal RelationsDouble Machine LearningFlexibilityMachine Learning Software
DoubleML is an open-source Python library implementing the double machine\nlearning framework of Chernozhukov et al. (2018) for a variety of causal\nmodels. It contains functionalities for valid statistical inference on causal\nparameters when the estimation of nuisance parameters is based on machine\nlearning methods. The object-oriented implementation of DoubleML provides a\nhigh flexibility in terms of model specifications and makes it easily\nextendable. The package is distributed under the MIT license and relies on core\nlibraries from the scientific Python ecosystem: scikit-learn, numpy, pandas,\nscipy, statsmodels and joblib. Source code, documentation and an extensive user\nguide can be found at https://github.com/DoubleML/doubleml-for-py and\nhttps://docs.doubleml.org.\n