Fairness-aware Learning through Regularization Approach

Abstract

Abstract-With the spread of data mining technologies and the accumulation of social data, such technologies and data are being used for determinations that seriously affect people's lives. For example, credit scoring is frequently determined based on the records of past credit data together with statistical prediction techniques. Needless to say, such determinations must be socially and legally fair from a viewpoint of social responsibility; namely, it must be unbiased and nondiscriminatory in sensitive features, such as race, gender, religion, and so on. Several researchers have recently begun to attempt the development of analysis techniques that are aware of social fairness or discrimination. They have shown that simply avoiding the use of sensitive features is insufficient for eliminating biases in determinations, due to the indirect influence of sensitive information. From a privacy-preserving viewpoint, this can be interpreted as hiding sensitive information when classification results are observed. In this paper, we first discuss three causes of unfairness in machine learning. We then propose a regularization approach that is applicable to any prediction algorithm with probabilistic discriminative models. We further apply this approach to logistic regression and empirically show its effectiveness and efficiency.

Fairness-Aware Classifier with Prejudice Remover Regularizer

“…Note that in our preliminary work [12], we took the approach of replacing X withx s , which is a sample mean vector of x over a set of training samples whose corresponding sensitive feature is equal to s. However, we unfortunately failed to obtain good approximations by this approach. Finally, the prejudice remover regularizer R PR (D, Θ) is…”

“…Note that we slightly changed this algorithm as described in [12], because the original algorithm may fail to stop.…”

Anti-discrimination learning: a causal modeling-based framework

“…In [12], the distribution of the non-protected attributes in the dataset is modified such that the protected attribute cannot be estimated from the nonprotected attributes. Proposed methods for discrimination prevention using algorithm tweaking require some tweak of predictive models [7,9,14,15,18,19,35]. For example, in [18], the authors developed a strategy for relabeling the leaf nodes of a decision tree to make it discrimination-free.…”

“…These approaches classify discrimination into different types such as group discrimination, individual discrimination, direct and indirect discrimination. Based on that, methods for discrimination prevention have been proposed [1,7,9,[12][13][14][15][17][18][19]21,35,37,43] which either use data preprocessing or algorithm tweaking. However, these works are mainly based on correlation or association-based measures which cannot be used to estimate the causal effect of the protected attributes on the decision.…”

Decision Theory for Discrimination-Aware Classification

“…Proposed methods for discrimination prevention using model adaptation include the tweaking of decision trees [2], naive Bayes classifiers [1], and logistic regression [8]. All these methods require that the learning model or algorithm is tweaked, and the first two methods are specific to their respective classifiers.…”