The paper argues that less discriminatory algorithms cannot be defined by quantitative performance metrics alone and must incorporate a reasonableness standard, with feasible but computationally hard search problems.
Model Selection's Disparate Impact in Real-World Deep Learning Applications
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
Algorithmic fairness has emphasized the role of biased data in automated decision outcomes. Recently, there has been a shift in attention to sources of bias that implicate fairness in other stages in the ML pipeline. We contend that one source of such bias, human preferences in model selection, remains under-explored in terms of its role in disparate impact across demographic groups. Using a deep learning model trained on real-world medical imaging data, we verify our claim empirically and argue that choice of metric for model comparison, especially those that do not take variability into account, can significantly bias model selection outcomes.
fields
cs.CY 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
What Constitutes a Less Discriminatory Algorithm?
The paper argues that less discriminatory algorithms cannot be defined by quantitative performance metrics alone and must incorporate a reasonableness standard, with feasible but computationally hard search problems.