Materials for "Counterfactual Evaluation and Learning for Interactive Systems: Foundations, Implementations, and Recent Advances", a tutorial delivered at the SIGKDD Conference on Knowledge Discovery and Data Mining (KDD'22).
-
Presenters: Yuta Saito (Cornell University, USA) and Thorsten Joachims (Cornell University, USA).
-
Tutorial website: https://counterfactual-ml.github.io/kdd2022-tutorial/
-
Tutorial proposal: https://dl.acm.org/doi/10.1145/3534678.3542601
-
Reference Package (Open Bandit Pipeline): https://github.com/st-tech/zr-obp
-
Survey of related papers: https://github.com/hanjuku-kaso/awesome-offline-rl
- examples: brief examples describing how to use Open Bandit Pipeline with synthetic data, classification data, and real-world bandit data
- simulations: simulation codes comparing a wide variety of existing OPE estimators on synthetic data
- real-world: a brief demo of OPE/OPL on real bandit dataset (need Open Bandit Dataset)
The Python environment is built using poetry. You can build the same environment as in our examples and simulations by cloning the repository and running poetry install directly under the folder (if you have not install poetry yet, please run pip install poetry first.).
# clone the obp repository
git clone https://github.com/usaito/kdd2022-tutorial.git
cd kdd2022-tutorial
# build the environment with poetry
poetry install
# activate jupyter-lab environment
poetry run jupyter lab
The versions of Python and used packages are as follows.
[tool.poetry.dependencies]
python = "^3.9,<3.10"
obp = "^0.5.4"
numpy = "^1.22.3"
matplotlib = "^3.5.2"
If you have any question, please feel free to contact: ys552@cornell.edu