Classification with no discrimination by preferential sampling

By Faisal Kamiran, Toon Calders. 2010

184 citations recorded by the source corpus.

Subfields: Evaluation & Benchmarks, AI Ethics, Fairness & Society

How researchers see this paper

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

In Evaluation & Benchmarks

In AI Ethics, Fairness & Society

Citations in the corpus

A Survey on Bias and Fairness in Machine Learning (2019, methodology): “Area Reference(s) Classification [78] [106] [57] [85] [147] [63] [159] [154] [69] [25] [155] [122] [49] [73] [75] Regression [14] [1] PCA [137] Community detection [104] Clustering [31] [8] Graph embedding [22] Causal inference [96] [164] [165] [160] [116] [115] [162] [82] [127] [161] Variational au”
Taking Advantage of Multitask Learning for Fair Classification (2018, background): “This is in contrast to other approaches which impose the fairness constraint as a post-processing step [10 19 20 31] or by modifying the data representation before employing standard machine learning methods [1 12 22–24 39].”
Fairness Beyond Disparate Treatment & Disparate Impact: Learning Classification without Disparate Mistreatment (2016, methodology): “These prior designs have attempted to tackle unfairness in decision making scenarios where the historical decisions in the training data are biased ( i.e. , groups of people with certain sensitive attributes may have historically received unfair treatment) and there is no ground truth about the…”
Three naive Bayes approaches for discrimination-free classification (2010, background): “Other existing approaches include resampling [7], and the construction of discrimination-aware decision trees [2].”
Learning Fair Classifiers (2015, methodology): “We first experiment with our first formulation and show that it allows for fine-grained fairness control on several wellknown classifiers, often at a minimal loss in terms of accuracy, and provides more flexibility than alternatives (Kamiran & Calders, 2010; Kamishima et al., 2011).”
Censoring Representations with an Adversary (2015, methodology): “Another approach is data massaging , whereby the labels of the training data are changed so that the training data is fair (Kamiran & Calders (2009)), this is similar to the resampling methods often used to tackle class-imbalance problems.”
From Parity to Preference-based Notions of Fairness in Classification (2017, background): “A flurry of recent research on fair learning has focused on defining appropriate notions of fairness and then designing mechanisms to ensure fairness in automated decision making [12, 14, 18, 19, 20, 21, 28, 32, 33, 34].”
The ethics of algorithms: Mapping the debate (2016, neutral): “…may be possible to direct algorithms not to consider sensitive attributes that contribute to discrimination (Barocas and Selbst, 2015), such as gender or ethnicity (Calders et al., 2009; Kamiran and Calders, 2010; Schermer, 2011), based upon the emergence of discrimination in a particular context.”
Empirical Risk Minimization under Fairness Constraints (2018, methodology): “Finally, the third family of methods implements fairness by modifying the data representation and then employs standard machine learning methods: [31, 7, 32, 4, 33, 34] are examples of these methods but, again, no consistency property nor comparison with state-of-the-art proposal are provided.”
Using Loglinear Model for Discrimination Discovery and Prevention (2016, background): “Previous works of discrimination prevention tried to address this challenging problem by introducing a reverse discrimination in the training data [26] or pushing constraints into the trained classifiers [7], [8].”
Handling Conditional Discrimination (2011, methodology): “The previous works solve the problem by introducing a reverse discrimination either in the training data [2], [5] or pushing constraints into the trained classifiers [3], [4].”
Explainable and Non-explainable Discrimination in Classification (2013, background): “In the existing discrimination-aware classification (Calders et al., 2009; Calders and Verwer, 2010; Kamiran et al., 2010; Kamiran and Calders, 2010) the discrimination is considered to be present if the acceptance rate for the favored community (denote m for males) and the deprived community…”
Beyond Distributive Fairness in Algorithmic Decision Making: Feature Selection for Procedurally Fair Learning (2018, background): “These concerns have spawned much recent research on detecting and avoiding unfairness in decision making (Dwork et al. 2012; Feldman et al. 2015; Kamiran and Calders 2010; Luong, Ruggieri, and Turini 2011; Pedreschi, Ruggieri, and Turini 2008; Zafar et al. 2017b; Zemel et al. 2013).”
Fairness Constraints: Mechanisms for Fair Classification (2015, methodology): “These studies have typically adopted one of the two following strategies: The first strategy consists of pre-processing the training data [Dwork et al., 2012, Feldman et al., 2015, Kamiran and Calders, 2009, 2010].”
Discrimination-aware classification (2011, methodology): “The discrimination-aware techniques discussed in this chapter are published in: IEEE conference on computer, control and communication [46]; Benelux conference on artificial intelligence [47]; the annual machine learning conference of Belgium and The Netherlands [48], and domain driven data mining w”
Rényi Fair Inference (2019, methodology): “Pre-processing methods modify the training data to remove discriminatory information before passing data to the decision-making process Calders et al. (2009); Feldman et al. (2015); Kamiran & Calders (2010; 2009; 2012); Dwork et al. (2012); Calmon et al. (2017); Ruggieri (2014).”
A survey on datasets for fairness‐aware machine learning (2021, unknown): “We discard the attribute fnlwgt (ﬁnal weight) as the suggestions of related work (B. H. Zhang, Lemoine, & Mitchell, 2018; Kamiran & Calders, 2012; Calders, Kamiran, & Pechenizkiy, 2009; Calders & Kamiran, 2010).”
Discrimination Aware Decision Tree Learning (2010, unknown): “Different forms of discrimination in employment, income, education, ﬁnance and in many other social activities may be based on age, gender, skin color, religion, race, language, culture, marital status, economic condition etc.”