Learning Decision Policies with Instrumental Variables through Double Machine Learning

Authors: Daqian Shao, Ashkan Soleymani, Francesco Quinzan, Marta Kwiatkowska

ICML 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this section, we empirically evaluate DML-IV for IV regression and offline IV bandit problems.
Researcher Affiliation Academia 1Department of Computer Science, University of Oxford, UK 2Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, USA.
Pseudocode Yes Algorithm 1 DML-IV with K-fold cross-fitting
Open Source Code Yes The algorithms are implemented using Py Torch (Paszke et al., 2019), and the code is available on Git Hub3. 3https://github.com/shaodaqian/DML-IV
Open Datasets Yes We consider two semi-synthetic real-world datasets IHDP4 (Hill, 2011) and PM-CMR5 (Wyatt et al., 2020).
Dataset Splits Yes we randomly split them into training (63%), validation (27%), and testing (10%) following Wu et al. (2023).
Hardware Specification No The paper does not specify the hardware used for the experiments (e.g., GPU models, CPU types, or memory specifications).
Software Dependencies No The algorithms are implemented using Py Torch (Paszke et al., 2019), and the code is available on Git Hub3. While PyTorch is mentioned, a specific version number is not provided, nor are other software dependencies with versions.
Experiment Setup Yes In this section we use DNN estimators for both stages with network architecture and hyper-parameters provided in Appendix F. Additional results of DML-IV using tree-based estimators such as Random Forests and Gradient Boosting are provided in Appendix G.2, where SOTA performance is also demonstrated.