Analyzing the impact of missing values and selection bias on fairness

Analyzing the impact of missing values and selection bias on fairness
复制标题

分析缺失值和选择偏差对公平性的影响

DOI:
10.1007/s41060-021-00259-z
复制
发表时间:
2021
影响因子:
2.4
通讯作者:
Singh, Lisa
Singh, Lisa
中科院分区:
--
文献类型:
--
作者:
Wang, Yanchen;Singh, Lisa

文献摘要

参考文献

被引文献

相似文献

算法决策变得越来越普遍,越来越多地影响着人们的日常生活。最近,关于机器决策公平性的讨论不断出现。研究人员提出了不同的方法来提高这些算法的公平性。虽然这些方法可以帮助机器做出更公平的决策,但它们是在相当干净的数据集上开发和验证的。不幸的是,大多数真实世界的数据都具有复杂性,这使得它们更脏。这项工作通过分析两个现实世界的数据问题对分类数据的公平性——缺失值和选择偏差的影响,考虑了其中的两个复杂性。在阐述了这个问题并表明其存在之后,我们提出了针对包含缺失值和/或选择偏差的数据集的修复算法,这些算法使用基于缺失值生成过程的不同形式的重加权和重采样。我们使用各种公平性指标对真实世界和合成数据进行了广泛的实证评估,并展示了不同机制和选择偏差产生的不同缺失值如何影响预测公平性,即使预测精度保持相当恒定。
Algorithmic decision making is becoming more prevalent, increasingly impacting people’s daily lives. Recently, discussions have been emerging about the fairness of decisions made by machines. Researchers have proposed different approaches for improving the fairness of these algorithms. While these approaches can help machines make fairer decisions, they have been developed and validated on fairlycleandata sets. Unfortunately, most real-world data have complexities that make them moredirty. This work considers two of these complexities by analyzing the impact of two real-world data issues on fairness—missing values and selection bias—for categorical data. After formulating this problem and showing its existence, we propose fixing algorithms for data sets containing missing values and/or selection bias that use different forms of reweighting and resampling based upon the missing value generation process. We conduct an extensive empirical evaluation on both real-world and synthetic data using various fairness metrics, and demonstrate how different missing values generated from different mechanisms and selection bias impact prediction fairness, even when prediction accuracy remains fairly constant.
DOI: 10.1186/s12874-017-0366-9
发表时间: 2017-07-04
影响因子: 4
作者:
Jahn-Eimermacher A;Ingel K;Preussler S;Bayes-Genis A;Binder H
通讯作者: Binder H
DOI: 10.1016/s0022-5347(17)55253-2
发表时间: 2015
期刊: Masui. The Japanese journal of anesthesiology
影响因子: --
作者:
S. Sakaue;M. Sunagawa;H. Tanigawa;Yu Saito;Shi;M. Okada;Akiou Nakamura;K. Arai;T. Hisamitsu
通讯作者: T. Hisamitsu
DOI: 10.1186/s12874-017-0442-1
发表时间: 2017-12-06
影响因子: 4
作者:
Jakobsen JC;Gluud C;Wetterslev J;Winkel P
通讯作者: Winkel P
数据缺失:问题与解决方案
DOI: --
发表时间: 2003
期刊:
影响因子: --
作者:
A. Pickles
通讯作者: A. Pickles
不利影响和测试验证:有效且合理的就业测试的从业者指南
DOI: --
发表时间: 2005
期刊:
影响因子: --
作者:
Dan A. Biddle
通讯作者: Dan A. Biddle