Causal Fairness Analysis - Software Tools
This page contains a sequence of vignettes associated with the Causal Fairness Analysis paper (Plečko and Bareinboim 2024). In particular, here you can find all the code used to reproduce the results of the paper, formatted in a vignette-style fashion. There are also additional examples of analyses, which are not presented in the manuscript. In particular, you can find the following examples, grouped by different tasks (bias detection, fair prediction, fair decision-making):
| Task | Dataset | Vignette Name | Manuscript Reference |
|---|---|---|---|
| Task 1 | Census 2018 | Census Bias Detection | Example 5.1, Section 5.1 |
| Task 1 | College Admissions | Admissions Over Time | Example 5.2, Section 5.1 |
| Task 1 | COMPAS | True Outcome vs. Northpointe’s Predictions | Example 5.3, Section 5.1 |
| Task 1 | COMPAS | Task 1 Beyond the SFM | Example 6.1, Section 6 |
| Task 1 | MIMIC-IV | Mortality after ICU admission | |
| Task 2 | COMPAS | Fair Prediction Theorem in the Wild | Example 5.4, Section 5.2 |
| Task 2 | COMPAS | Neural In-Processing | |
| Task 3 | MIMIC-IV | Respirators in the ICU | |
| Task 3 | Surgeries (Synthetic) | Cancer Surgery Allocation | Section 5.3.3 |
| Other | Adult / COMPAS | Conditional independence tests | Figure 4.7, Section 4.5 |
| Other | COMPAS | Using CFA from python |
Want to learn more about Causal Fairness Analysis?
For those interested in learning more about CFA, we suggest the following resources:
- Reading the Causal Fairness Analysis paper, found here,
- Follow the series of lectures on CFA which were part of the COMSW-4775 course at Columbia Computer Science,
- Check our ICML 2022 Tutorial.
- Check the vignettes on this webpage which demonstrate how to perform Causal Fairness Analysis in practice.
References
Plečko, Drago, and Elias Bareinboim. 2024. “Causal Fairness Analysis: A Causal Toolkit for Fair Machine Learning.” Foundations and Trends in Machine Learning 17 (3): 304–589.