The accuracy, fairness, and limits of predicting recidivism.

The accuracy, fairness, and limits of predicting recidivism.
复制标题

预测累犯的准确性,公平性和局限性。

DOI:
10.1126/sciadv.aao5580
复制
发表时间:
2018-01
期刊:
影响因子:
13.6
通讯作者:
Farid H
Farid H
中科院分区:
综合性期刊1区
文献类型:
--
作者:
Dressel J;Farid H

文献摘要

参考文献

被引文献

相似文献

我们应该相信计算机在刑事司法系统中做出改变生活的决定吗?预测累犯的算法通常被用来评估刑事被告犯罪的可能性。这些预测被用于预审、假释和量刑决定。这些系统的支持者认为,大数据和先进的机器学习使这些分析比人类更准确,偏见更少。然而,我们表明,广泛使用的商业风险评估软件COMPAS并不比几乎没有刑事司法专业知识的人所做的预测更准确或更公平。我们进一步证明,一个只有两个特征的简单线性预测器与具有137个特征的COMPAS几乎是等价的。
Should we trust computers to make life-altering decisions in the criminal justice system? Algorithms for predicting recidivism are commonly used to assess a criminal defendant’s likelihood of committing a crime. These predictions are used in pretrial, parole, and sentencing decisions. Proponents of these systems argue that big data and advanced machine learning make these analyses more accurate and less biased than humans. We show, however, that the widely used commercial risk assessment software COMPAS is no more accurate or fair than predictions made by people with little or no criminal justice expertise. We further show that a simple linear predictor provided with only two features is nearly equivalent to COMPAS with its 137 features.
DOI: 10.1111/j.1745-9125.1996.tb01220.x
发表时间: 1996-11-01
期刊: CRIMINOLOGY
影响因子: 5.8
作者:
Gendreau, P;Little, T;Goggin, C
通讯作者: Goggin, C
DOI: 10.1177/0093854800027001002
发表时间: 2000-02-01
影响因子: 2.1
作者:
Hanson, RK;Harris, AJR
通讯作者: Harris, AJR
DOI: 10.1177/107906320201400206
发表时间: 2002-04-01
期刊: Sexual abuse : a journal of research and treatment
影响因子: --
作者:
Beech, Anthony;Friendship, Caroline;Hanson, R Karl
通讯作者: Hanson, R Karl
DOI: 10.1037/0033-295x.112.2.494
发表时间: 2005-04-01
影响因子: 5.4
作者:
Hastie, R;Kameda, T
通讯作者: Kameda, T
DOI: 10.1037/a0020473
发表时间: 2010-09-01
影响因子: 22.4
作者:
Yang, Min;Wong, Stephen C. P.;Coid, Jeremy
通讯作者: Coid, Jeremy