Robust Fair Clustering: A Novel Fairness Attack and Defense Framework
Authors: Anshuman Chhabra, Peizhao Li, Prasant Mohapatra, Hongfu Liu
ICLR 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Through comprehensive experiments1, we find that state-of-the-art models are highly susceptible to our attack as it can reduce their fairness performance significantly. Finally, we propose Consensus Fair Clustering (CFC), the first robust fair clustering approach that transforms consensus clustering into a fair graph partitioning problem, and iteratively learns to generate fair cluster outputs. Experimentally, we observe that CFC is highly robust to the proposed attack and is thus a truly robust fair clustering alternative. |
| Researcher Affiliation | Academia | Department of Computer Science, University of California, Davis {chhabra,pmohapatra}@ucdavis.edu Michtom School of Computer Science, Brandeis University {peizhaoli,hongfuliu}@brandeis.edu |
| Pseudocode | No | The paper describes the steps of the proposed methods (e.g., in Section 3.1 'Defense Algorithm to Fairness Attack') but does not provide structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | 1Code available here: https://github.com/anshuman23/CFC. |
| Open Datasets | Yes | We construct MNIST-USPS dataset using all the training digital samples from MNIST (Le Cun, 1998) and USPS dataset (Le Cun, 1990)... Office-31: The Office-31 dataset (Saenko et al., 2010)... Note that we also conduct experiments on the Inverted UCI DIGITS dataset (Xu et al., 1992) and Extended Yale Face B dataset (Lee et al., 2005)... |
| Dataset Splits | No | The paper mentions using 'all the training digital samples from MNIST' but does not provide specific training/validation/test dataset split percentages, absolute sample counts for each split, or detailed splitting methodology. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, processor types, memory amounts, or detailed computer specifications) used for running its experiments. |
| Software Dependencies | No | The paper mentions implementing algorithms in Python (Appendix A) and using specific algorithms like K-means (Lloyd, 1982), but does not provide specific version numbers for Python or any other software libraries or dependencies used in the experiments. |
| Experiment Setup | Yes | For SFD, we set the parameters p = 2, q = 5 for all datasets except DIGITS for which we set p = 1, q = 5. For FSC we use the default parameters and use the nearest neighbors approach Von Luxburg (2007) for creating the input graph for which we set the number of neighbors = 3 for all datasets. For KFC we use the default parameter value of δ = 0.1. The hyperparameters such as number of basic partitions r, temperature parameter τ in the contrastive loss Lc, dropout in hidden layers, number of training epochs, the activation function [...] These are r = 100, τ = 2, dropout = 0.6, # epochs = 3000, Gaussian Error Linear Unit (Hendrycks & Gimpel, 2016) is used as the activation function... Moreover, the dimension of the hidden layer is set to 256 for all datasets except for DIGITS since DIGITS has only 64 features and hence we use the hidden layer dimension as 36 for it. Using grid based search we set the following parameters for the given datasets: for Office-31 we have R = 1, α = 1, β = 100; for MNIST-USPS we have R = 2, α = 100, β = 25; for Yale we have R = 2, α = 50, β = 10; and for DIGITS we have R = 2, α = 10, β = 50. |