一句话：提出基于因果分析的自适应方法，为判别性神经网络选择最有效的公平性改进策略。
该论文研究神经网络公平性改进问题：给定一个存在歧视的判别性神经网络，目标是在不显著牺牲准确率的前提下降低歧视。已有预处理、处理中与后处理等公平性改进方法，但实证研究表明这些方法并不总是有效，可能以大幅牺牲准确率为代价，甚至同时恶化公平性与准确率。为此，作者提出一种基于因果分析的自适应方法，通过分析导致不公平的神经元与属性在输入属性及隐藏神经元之间的分布，据此选择最合适的公平性改进方法。实验评估表明，该方法能够始终识别出最佳的公平性改进方法，且效率较高，平均时间开销约为 5 分钟。
Given a discriminating neural network, the problem of fairness improvement is to systematically reduce discrimination without significantly scarifies its performance (i.e., accuracy). Multiple categories of fairness improving methods have been proposed for neural networks, including pre-processing, in-processing and post-processing. Our empirical study however shows that these methods are not always effective (e.g., they may improve fairness by paying the price of huge accuracy drop) or even not helpful (e.g., they may even worsen both fairness and accuracy). In this work, we propose an approach which adaptively chooses the fairness improving method based on causality analysis. That is, we choose the method based on how the neurons and attributes responsible for unfairness are distributed among the input attributes and the hidden neurons. Our experimental evaluation shows that our approach is effective (i.e., always identify the best fairness improving method) and efficient (i.e., with an average time overhead of 5 minutes).
DOI 原文 ·
@inproceedings{paperbot7004,
title = {Adaptive fairness improvement based on causality analysis},
author = {Mengdi Zhang and Jun Sun},
booktitle = {FSE},
year = {2022},
doi = {10.1145/3540250.3549103}
}