We present an empirical study of debiasing methods for classifiers, showing that debiasers often fail in practice to generalize out-of-sample, and can in fact make fairness worse rather than better.
We study the effects of different debiasing methods on the underlying causal path specific effects (PSEs) and explain these effects using an information-theoretic perspective.
We present Fairness.jl, a comprehensive bias audit and mitigation toolkit in julia. Extensive support and functionality provided by MLJ.jl has been used in this package.
Many methods for debiasing classifiers have been proposed, but their effectiveness in practice remains unclear. We evaluate the performance of pre-processing and post-processing debiasers for improving fairness in random forest classifiers trained on …