A multidisciplinary survey on discrimination analysis

By Andrea Romei, Salvatore Ruggieri. 2013

279 citations recorded by the source corpus.

Subfields: Economics and Econometrics, Marketing, Sociology and Political Science, Statistics and Probability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

In Marketing

In Sociology and Political Science

In Statistics and Probability

Citations in the corpus

A Survey on Bias and Fairness in Machine Learning (2019, background): “[99, 133, 152] contain more broad information on discrimination theory that involve more multidisciplinary concepts from legal theory, economics, and social sciences which can be referenced by the interested readers.”
Fair decision making using privacy-protected data (2019, background): “a formal model for reasoning about and controlling a quantitative measure of privacy loss. Fairness has been formalized in economics, and, more recently, in definitions emerging from machine learning [7, 8, 19, 22, 26]. Yet relatively little work has considered the direct interaction of privacy and ”
A Unified Approach to Quantifying Algorithmic Unfairness: Measuring Individual &Group Unfairness via Inequality Indices (2018, background): “As algorithmic decision making systems are increasingly used in life-affecting scenarios such as criminal risk prediction [2, 6] and credit risk assessments [15], concerns have risen about the potential unfairness of these decisions to certain social groups or individuals [4, 32, 33].”
A Survey of Methods for Explaining Black Box Models (2018, background): “The first principle requires that the model guarantees the protection of groups against (direct or indirect) discrimination [100]; while the second one requires that the model does not reveal sensitive information about people [4].”
On formalizing fairness in prediction with machine learning (2017, methodology): “For a multidisciplinary survey on discrimination analysis, we recommend Romei and Ruggieri [58].”
Conscientious Classification: A Data Scientist's Guide to Discrimination-Aware Classification (2017, positive): “9,17 We present in this section several high-level methods following the organization nomenclature of Hajian et al. from their tutorial in the 2016 ACMSigKDD conference.”
Iterative Orthogonal Feature Projection for Diagnosing Bias in Black-Box Models (2016, background): “The literature on fairness is broad including works from social choice theory, game theory, economics, and law (Romei & Ruggieri, 2014).”
Fair Prediction with Disparate Impact: A Study of Bias in Recidivism Prediction Instruments (2016, neutral): “4 2.2 Further related work Though the study of discrimination in decision making and predictive modeling is rapidly evolving, it also has a long and rich multidisciplinary history. Romei and Ruggieri [18] provide an excellent overview of some of the work in this broad subject area. The recent work o”
Auditing black-box models for indirect influence (2016, background): “Finally, we note that a compelling application for blackbox audits of indirect influence includes considerations of algorithmic fairness [22].”
Fairness-aware Classification: Criterion, Convexity, and Bounds (2018, background): “Risk ratio is a common fairness notion (Pedreschi, Ruggieri, and Turini 2009; Romei and Ruggieri 2014).”
On Fairness and Calibration (2017, background): “(More thorough reviews can be found in [2, 4, 32]).”
Algorithmic Decision Making and the Cost of Fairness (2017, background): “f COMPAS is that the rate of false positives is higher among blacks than whites [2]. 2.2 Related work „e literature on designing fair algorithms is extensive and interdisciplinary. Romei and Ruggieri [36] and Zliobaite [42] survey various measures of fairness in decision making. Here we focus on alg”
A Causal Framework for Discovering and Removing Direct and Indirect Discrimination (2016, background): “…the total causal e ﬀ ect, i.e., It can be proved straightforwardly using Equation (2) that the above equation equals to P ( e + | c + ) − P ( e + | c − ), which is known as risk di ﬀ erence (Romei and Ruggieri 2014) widely used for discrimination measurement in the anti-discrimination literature.”
Fairness Beyond Disparate Treatment & Disparate Impact: Learning Classification without Disparate Mistreatment (2016, methodology): “Finally, we experiment with synthetic as well as real world datasets and show that our methodology can be eﬀectively used to avoid disparate mistreatment.”
A survey on measuring indirect discrimination in machine learning (2015, background): “An interested reader is referred to focused surveys [Romei and Ruggieri 2014; Barocas and Selbst 2016] for more information.”
Learning Fair Classifiers (2015, background): “…of a decision making process (Barocas & Selbst, 2016; Biddle, 2005): (i) disparate treatment: the decision making process is (partly) based on information about sensitive attributes (e.g., gender, race) of subjects; and, (ii) disparate impact: the outcomes of the decision making process have a…”