On the Fairness of Causal Algorithmic Recourse

Authors: Julius von Kügelgen, Amir-Hossein Karimi, Umang Bhatt, Isabel Valera, Adrian Weller, Bernhard Schölkopf9584-9594

AAAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We study theoretically and empirically how to enforce fair causal recourse by altering the classifier and perform a case study on the Adult dataset.
Researcher Affiliation Academia 1 Max Planck Institute for Intelligent Systems, T ubingen, Germany 2 University of Cambridge 3 ETH Z urich 4 Saarland University 5 The Alan Turing Institute
Pseudocode No The paper describes optimization problems (e.g., Eq. 1, 2, 3) and solution approaches (e.g., 'using brute force search'), but it does not include any explicitly labeled 'Pseudocode' or 'Algorithm' blocks.
Open Source Code Yes Code to reproduce our experiments is available at https://github.com/amirhk/recourse.
Open Datasets Yes We use the Adult dataset (Lichman et al. 2013), which consists of 45k+ samples without missing data.
Dataset Splits Yes select all remaining hyperparameters (including the trade-off parameter λ for Fair SVM) using 5-fold cross validation.
Hardware Specification No The paper describes numerical simulations and a case study on the Adult dataset but does not provide specific details regarding the hardware used for these experiments.
Software Dependencies No The paper mentions training "linear and nonlinear logistic regression (LR), and different support vector machines (SVMs)" and using a "probabilistic framework of Karimi et al. (2020c)", but it does not specify exact version numbers for any software libraries or dependencies.
Experiment Setup Yes We use either a linear or polynomial kernel for all SVMs (depending on the GT labels) and select all remaining hyperparameters (including the trade-off parameter λ for Fair SVM) using 5-fold cross validation.