Stable and Fair Classification

Authors: Lingxiao Huang, Nisheeth Vishnoi

ICML 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We assess the benefits of our approach empirically by extending several fair classification algorithms that are shown to achieve a good balance between fairness and accuracy over the Adult dataset, and show that our framework improves the stability at only a slight sacrifice in accuracy.
Researcher Affiliation Academia 1EPFL, Switzerland 2Yale University, USA.
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide an explicit statement or link for the open-sourcing of its code.
Open Datasets Yes Our simulations are over an income dataset Adult (Dheeru & Karra Taniskidou, 2017), that records the demographics of 45222 individuals, along with a binary label indicating whether the income of an individual is greater than 50k USD or not. We use the pre-processed dataset as in (Friedler et al., 2019).
Dataset Splits Yes We perform 50 repetitions, in which we uniformly sample a training set (75%) from the remaining data.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers (e.g., Python 3.8, PyTorch 1.9).
Experiment Setup Yes For all three algorithms, we set the regularization parameter λ to be 0, 0.01, 0.02, 0.03, 0.04, 0.05 and compute the resulting stability metric stab, average accuracy and average fairness.