Identifying Representational Biases in Datasets Using PCA: A Max-Disparity Partition Framework
A greedy local-search algorithm is proposed, grounded in the Fiduccia-Mattheyses bipartitioning framework, that discovers the disparity-maximising partition without any predefined group labels and is passed directly to Fair PCA, completing a detect-explain-mitigate pipeline.
K. Arjun, Shashi Jain
· 0 citations