arXiv research

Study shows significant differences in recommendation bias between model-based and memory-based algorithms.

DCEM algorithm reduces bias in machine learning models trained on selective labels.

Study shows racial bias in health data, which can be reduced with simple techniques.

Study shows explanation disparities in machine learning models are influenced by data and model properties.

Deep chest X-ray classifiers show bias in predicting diagnoses.

What does it mean for an algorithm to be biased? In U.S. law, unintentional bias is encoded via disparate impact, which occurs when a selection process has widely different outcomes for different groups, even as it appears to be neutral. This legal determination hinges on a definition of a protected class (ethnicity, g…

Study reveals class disparities in balanced datasets through spectral imbalance.

Study highlights fairness issues in travel behavior prediction models.

New method reduces privacy impact on model accuracy for underrepresented groups.

Assessing the fairness of a decision making system with respect to a protected class, such as gender or race, is challenging when class membership labels are unavailable. Probabilistic models for predicting the protected class based on observable proxies, such as surname and geolocation for race, are sometimes used to …

Paper tackles fairness in CCA by minimizing correlation disparity error.

Bayesian model identifies health disparities in disease progression.

The paper examines bias in ML models using the Adult dataset.

EL framework certifies and flags bias in ML models without distributional assumptions.

Recidivism prediction instruments provide decision makers with an assessment of the likelihood that a criminal defendant will reoffend at a future point in time. While such instruments are gaining increasing popularity across the country, their use is attracting tremendous controversy. Much of the controversy concerns …

Study finds racial bias in pulse oximeter readings has minimal impact on ICU ventilation rates.

Recidivism prediction instruments (RPI's) provide decision makers with an assessment of the likelihood that a criminal defendant will reoffend at a future point in time. While such instruments are gaining increasing popularity across the country, their use is attracting tremendous controversy. Much of the controversy c…

Whereas previous post-processing approaches for increasing the fairness of predictions of biased classifiers address only group fairness, we propose a method for increasing both individual and group fairness. Our novel framework includes an individual bias detector used to prioritize data samples in a bias mitigation a…

Study federates measurement of demographic disparities from quantile sketches.

Study shows how missing data from certain groups can unfairly bias risk models.

We study the phenomenon of bias amplification in classifiers, wherein a machine learning model learns to predict classes with a greater disparity than the underlying ground truth. We demonstrate that bias amplification can arise via an inductive bias in gradient descent methods that results in the overestimation of the…

New method to quantify feature contributions to disparity without access to decision-making model.

Develops causal framework for fair survival analysis in healthcare.

Algorithmic decision making process now affects many aspects of our lives. Standard tools for machine learning, such as classification and regression, are subject to the bias in data, and thus direct application of such off-the-shelf tools could lead to a specific group being unfairly discriminated. Removing sensitive …

New method detects bias in AI models that generate data.

The paper develops a theory explaining how machine learning models can amplify biases.

Predictive modeling is increasingly being employed to assist human decision-makers. One purported advantage of replacing human judgment with computer models in high stakes settings-- such as sentencing, hiring, policing, college admissions, and parole decisions-- is the perceived "neutrality" of computers. It is argued…

This work analyzes PINNs for advection-diffusion equations using NTK theory.

Study reveals AI skin cancer classifiers underperform for darker skin phototypes, advocating for fairness auditing.

The paper shows how demographic data can lead to biased predictions, proposing 'Affirmative Information' as a solution.

New clustering method considers causal fairness to avoid bias.

The paper tackles fairness in estimating graphical models, especially for protected attributes.

Gradient clipping helps private SGD converge despite potential bias.

New models reduce regional inequality by adjusting exchange range and asset distribution bias.

Develops a framework for fair semi-supervised learning.

Language is increasingly being used to define rich visual recognition problems with supporting image collections sourced from the web. Structured prediction models are used in these tasks to take advantage of correlations between co-occurring labels and visual input but risk inadvertently encoding social biases found i…

New fair regression method improves fairness in chronic kidney disease classification.

New framework for interpreting disaggregated fairness evaluations using causal models.

The paper addresses bias amplification in prediction and decision-making using causal analysis.

The paper studies and mitigates accuracy disparity in regression models.

A hybrid strategy forecasts short-term loads using Warm-start Gradient Tree Boosting.

BIND removes background noise from binary matrices, improving detection accuracy and fairness.

Develops methods for fair classification under linear disparity constraints.

A membership inference attack (MIA) against a machine-learning model enables an attacker to determine whether a given data record was part of the model's training data or not. In this paper, we provide an in-depth study of the phenomenon of disparate vulnerability against MIAs: unequal success rate of MIAs against diff…

Following related work in law and policy, two notions of disparity have come to shape the study of fairness in algorithmic decision-making. Algorithms exhibit treatment disparity if they formally treat members of protected subgroups differently; algorithms exhibit impact disparity when outcomes differ across subgroups,…

Study decomposes racial healthcare disparities via shifts in mediator distributions.

Study shows label errors impact model disparity metrics, proposing mitigation methods.

Active learning has long been a topic of study in machine learning. However, as increasingly complex and opaque models have become standard practice, the process of active learning, too, has become more opaque. There has been little investigation into interpreting what specific trends and patterns an active learning st…