Guarantees for Spectral Clustering with Fairness Constraints

Authors: Matthäus Kleindessner, Samira Samadi, Pranjal Awasthi, Jamie Morgenstern

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

Reproducibility Variable Result LLM Response
Research Type Experimental In this section, we present a number of experiments. We first study our fair versions of spectral clustering, Algorithm 2 and Algorithm 3, on synthetic data generated according to our variant of the SBM and compare our algorithms to standard SC. We also study how robust our algorithms are with respect to a certain perturbation of our model. We then compare our algorithms to standard SC on real network data.
Researcher Affiliation Academia 1Department of Computer Science, Rutgers University, NJ 2College of Computing, Georgia Tech, GA.
Pseudocode Yes Algorithm 1 Unnormalized SC, Algorithm 2 Unnormalized SC with fairness constraints
Open Source Code Yes The code is available on https://github.com/matthklein/fair-spectral-clustering.
Open Datasets Yes The first row of Figure 5 shows the results as a function of the number of clusters k for two high school friendship networks (Mastrandrea et al., 2015)... The second row shows the results for DRUGNET, a network encoding acquaintanceship between drug users in Hartford, CT (Weeks et al., 2002).
Dataset Splits No The paper does not provide specific train/validation/test dataset splits for their experiments.
Hardware Specification No The paper does not provide specific hardware details (e.g., CPU, GPU models, memory) used for running the experiments.
Software Dependencies No We implemented all algorithms in MATLAB. We used the built-in function for k-means clustering... (No version numbers provided for MATLAB or specific libraries).
Experiment Setup Yes We used the built-in function for k-means clustering with all parameters set to their default values except for the number of replicates, which we set to 10.