Path-Specific Counterfactual Fairness

被引:0
作者
Chiappa, Silvia [1 ]
机构
[1] DeepMind London, London, England
来源
THIRTY-THIRD AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTY-FIRST INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE / NINTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE | 2019年
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We consider the problem of learning fair decision systems from data in which a sensitive attribute might affect the decision along both fair and unfair pathways. We introduce a counterfactual approach to disregard effects along unfair pathways that does not incur in the same loss of individual-specific information as previous approaches. Our method corrects observations adversely affected by the sensitive attribute, and uses these to form a decision. We leverage recent developments in deep learning and approximate inference to develop a VAE-type method that is widely applicable to complex nonlinear models.
引用
收藏
页码:7801 / 7808
页数:8
相关论文
共 24 条
[1]  
[Anonymous], DISCRETION HIRING
[2]  
[Anonymous], 2017, ADV NEURAL INFORM PR
[3]  
[Anonymous], 4 INT C LEARN REPR
[4]  
[Anonymous], 2014, 2 INT C LEARN REP IC
[5]  
[Anonymous], 2017, ELEMENTS CAUSAL INFE
[6]  
[Anonymous], 32 AAAI C ART INT
[7]  
[Anonymous], INT S 2017 SOC SCI
[8]  
[Anonymous], 2016, 4 INT C LEARN REPR
[9]  
BONCHI F., 2017, Int J Data Sci Anal, V3, P1, DOI DOI 10.1007/S41060-016-0040-Z
[10]   Explicit-Duration Markov Switching Models [J].
Chiappa, Silvia .
FOUNDATIONS AND TRENDS IN MACHINE LEARNING, 2014, 7 (06) :803-886