Racial categories in machine learning

Abstract

Controversies around race and machine learning have sparked debate among computer scientists over how to design machine learning systems that guarantee fairness. These debates rarely engage with how racial identity is embedded in our social experience, making for sociological and psychological complexity. This complexity challenges the paradigm of considering fairness to be a formal property of supervised learning with respect to protected personal attributes. Racial identity is not simply a personal subjective quality. For people labeled "Black" it is an ascribed political category that has consequences for social differentiation embedded in systemic patterns of social inequality achieved through both social and spatial segregation. In the United States, racial classification can best be understood as a system of inherently unequal status categories that places whites as the most privileged category while signifying the Negro/black category as stigmatized. Social stigma is reinforced through the unequal distribution of societal rewards and goods along racial lines that is reinforced by state, corporate, and civic institutions and practices. This creates a dilemma for society and designers: be blind to racial group disparities and thereby reify racialized social inequality by no longer measuring systemic inequality, or be conscious of racial categories in a way that itself reifies race. We propose a third option. By preceding group fairness interventions with unsupervised learning to dynamically detect patterns of segregation, machine learning systems can mitigate the root cause of social disparities, social segregation and stratification, without further anchoring status categories of disadvantage.

Saving Face

“…Although such a design is derived from the naturally occurring labels of the crawled and referenced datasets, the selected groupings have inherent limitations. Unlike the Pilot Parliaments Benchmark from the Gender Shades study [6] where the intersectional groups are defined with respect to skin type, "ethnicity" is an attribute that is highly correlated but not deterministically linked to racial categories, which are themselves nebulous social constructs, encompassing individuals with a wide range of phenotypic features [3]. Similarly, binary gender labels are compatible with the format of commercial product outputs, but exclusionary of those not presenting in the stereotypical representations of each selected gender identity [29].…”

Ethical dilemmas posed by mobile health and machine learning in psychiatry research

“…40 In machine learning, fairness encompasses concerns about how data-driven approaches can reflect and perpetuate biases rooted in social inequality and discrimination. 41,42 A model's predictions can vary systematically across demographic groups if, for example, the data being sampled reflects societal inequalities (i.e. historical bias) or if the sampling methods result in the underrepresentation of certain groups (i.e.…”

Bias Remediation in Driver Drowsiness Detection Systems Using Generative Adversarial Networks

“…Their study showed that there are gaps when using three different semantic representations such as bag-of-words, Deep Recurrent Neural Networks (DRNN), and word embedding [38]. Benthall and Haynes have investigated supervised learning algorithms and revealed that they are exposed to racial bias because of the differentiation that is embedded in systematic patterns [39].…”