Counterfactual Cross-Validation

---

About

This repository accompanies the semi-synthetic simulations conducted in the paper "Counterfactual Cross-Validation: Stable Model Selection Procedure for Causal Inference Models" by Yuta Saito and Shota Yasui, which has been accepted by ICML2020.

If you find this code useful in your research then please cite:

@article{saito2019counterfactual,
  title={Counterfactual Cross-Validation: Effective Causal Model Selection from Observational Data},
  author={Saito, Yuta and Yasui, Shota},
  journal={arXiv preprint arXiv:1909.05299},
  year={2019}
}

Dependencies

• numpy==1.16.2
• pandas==0.24.2
• scikit-learn==0.20.3
• tensorflow==1.15.2
• plotly==3.10.0
• econml==0.4
• optuna==0.19.0

Running the code

To run the simulation with IHDP data, navigate to the src/ directory and run the command

python main.py\
  --iters 100\
  --n_trials 100\
  --alpha_list 100 50 10 5 1 0.5 0.1 0.05 0.01

This will run semi-synthetic experiments (model selection and hyper-parameter tuning) conducted in Section 5 of the paper.

Reference

1. Shalit, U., Johansson, F. D., and Sontag, D. "Estimating individual treatment effect: generalization bounds and algorithms," In Proceedings of the 34th International Conference on Machine Learning-Volume 70, pp. 3076–3085. JMLR. org, 2017. github