Study shows racial bias in health data, which can be reduced with simple techniques.
problem Racial bias in health indicators measured by the Medical Expenditure Panel Survey (MEPS).
method Used publicly available and nationally representative MEPS data to show bias in predictive models for care management.
result Racial bias can be significantly reduced using simple mitigation techniques.
COMPAS recidivism predictions show racial bias against African Americans, study finds.
problem Racial bias in recidivism prediction algorithms.
method Causal analysis using FACT, a fairness measure grounded in causal inference.
result COMPAS shows racial bias against African American defendants, robust to unmeasured confounding.
Reduces gender classification bias by learning race-invariant face representations.
problem Societal bias in gender recognition systems.
method Adversarially trained autoencoder model to learn race-invariant face representations.
result Achieved a significant drop of over 40% in racial bias surrogate metric with race invariant representations.
Develops a method to quantify racial bias in law enforcement systems.
problem Quantify racial bias in law enforcement systems considering criminality and multi-stage interactions.
method Multi-stage causal framework incorporating criminality.
result Identifies three canonical scenarios of racial bias in law enforcement.
Study finds racial bias in pulse oximeter readings has minimal impact on ICU ventilation rates.
problem Racial disparities in pulse oximeter readings affect clinical decisions in ICU settings.
method Causal inference using path-specific effects and doubly robust estimator.
result Minimal impact of racial discrepancies on invasive ventilation rates, but more pronounced on ventilation duration.
Recidivism prediction scores are used across the USA to determine sentencing and supervision for hundreds of thousands of inmates. One such generator of recidivism prediction scores is Northpointe's Correctional Offender Management Profiling for Alternative Sanctions (COMPAS) score, used in states like California and F…
Develops tools to audit ML models for bias and unfairness.
problem Auditing ML models for individual bias and unfairness.
method Formalizes the task as an optimization problem and develops inferential tools for the optimal value.
result Demonstrates the utility of tools in revealing biases in COMPAS recidivism prediction instrument.
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…
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 complexi…
Study decomposes racial healthcare disparities via shifts in mediator distributions.
problem Racial disparities in healthcare expenditures and their underlying drivers.
method Framework decomposing disparities into mediator distribution shifts and residual components, using MEPS data.
result Substantial disparities persist even when mediators are equalized, suggesting unmeasured or structural factors.
SLUG method detects bias and out-of-distribution content in generative models.
problem Generative models can underrepresent certain groups and fail on out-of-distribution data.
method SLUG: A new uncertainty quantification method for VAEs combining Laplace approximations and stochastic trace estimators.
result SLUG's UQ score correlates with bias and out-of-distribution content.
The paper examines how machine learning tools in justice settings can unfairly affect different racial groups.
problem Machine learning tools in justice settings can unfairly affect different racial groups.
method Exploring different ideas of racial equity and their computational trade-offs.
result Computation alone is unlikely to solve the unfairness in machine learning tools for justice settings.
This paper assesses biases in contextualized word representations.
problem Analyzing biases in contextualized word representations.
method Proposes assessing bias at the contextual word level, capturing contextual effects of bias.
result Demonstrates evidence of bias in contextual word models, including racial bias and exacerbated effects for intersectional minorities.
Study shows how algorithmic prediction affects US housing market, reducing racial wealth disparities.
problem Impact of algorithmic prediction on housing market and racial wealth disparities.
method Natural experiment using digitization of housing records to study entry, allocation, and prices.
result Digitization leads to increased sale prices for minority-owned homes, reducing racial wealth disparities.
The study uses transfer learning to compare surgical outcomes across racial/ethnic subgroups.
problem Difficulty in comparing surgical outcomes due to racial/ethnic and geographic differences.
method Causal inference framework and transfer learning to incorporate data from multiple populations.
result Racial and ethnic differences in surgical outcomes are found, with non-Hispanic Black patients experiencing wide variability.
Introduces privilege scores to measure and interpret protected attribute-related privilege in machine learning models.
problem Lack of explicit formulation of non-neutrality in fairness-aware machine learning methods.
method Privilege scores (PS) and privilege score contributions (PSCs) to measure and interpret protected attribute-related privilege.
result Demonstrates the broad applicability of PS and PSCs in gender and racial privilege in mortgage and college admissions applications.
Develops causal framework for fair survival analysis in healthcare.
problem Fairness in survival analysis for high-stakes domains like healthcare.
method Causal framework using graphical models, conditional survival function, and Causal Reduction Theorem.
result Decomposes disparities in survival into direct, indirect, and spurious pathways.
Method debiases alternative data for fair credit underwriting.
problem Bias in alternative data affecting credit underwriting fairness.
method Causal inference applied to machine learning models.
result Improves model accuracy across racial groups without discrimination.
Prior work finds a diversity paradox: diversity breeds innovation, and yet, underrepresented groups that diversify organizations have less successful careers within them. Does the diversity paradox hold for scientists as well? We study this by utilizing a near-population of ~1.2 million US doctoral recipients from 1977…
Study reveals AI skin cancer classifiers underperform for darker skin phototypes, advocating for fairness auditing.
problem AI bias in dermatology, particularly for darker skin phototypes.
method Predictive Representativity (PR) framework, evaluating classifiers on HAM10000 and BOSQUE Test sets.
result Substantial performance disparities by skin phototype, highlighting AI bias.
New method for analyzing compositional data, addressing biases in summary statistics.
problem Inadequate effect measures for compositional data, especially in high-dimensionality and sparsity.
method Perturbation-based effect measures, average perturbation effects.
result Proposed estimators efficiently estimate average perturbation effects, outperforming existing techniques.
The paper tackles individual fairness in ML models, developing statistical methods to detect bias.
problem Detecting and measuring violations of individual fairness in machine learning models.
method Formalizing the problem as adversarial attack, developing inference tools for the adversarial cost function.
result Statistical methods to assess and test hypotheses of model fairness with non-coverage error rate control.
The paper tackles multi-level fairness in algorithmic systems, addressing bias at both individual and structural levels.
problem Algorithmic systems can unfairly impact marginalized groups, especially when considering only individual-level bias.
method Formalizes multi-level fairness using causal inference tools, addressing effects of sensitive attributes at multiple levels.
result Illustrates the importance of accounting for macro-level sensitive attributes in fairness assessments.
MASC balances dataset representation using affinity clustering and distribution discrepancies.
problem Representation bias in datasets due to group imbalance.
method MASC uses affinity clustering and pairwise distribution discrepancies to balance non-protected and protected groups.
result MASC effectively debiases target datasets, comparable to existing methods.
Improved surname geocoding and name supplements enhance race imputation accuracy.
problem Census data problems affecting race imputation accuracy.
method Fully Bayesian Improved Surname Geocoding (fBISG) and name supplements.
result Significant improvement in race imputation accuracy across all racial groups.
Framework generates fair synthetic data to avoid biases.
problem Societal and historic biases in training data lead to biased algorithms.
method Self-supervised learning with fairness constraints.
result Generated fair synthetic data maintains relationships while controlling biases.
To answer questions about racial inequality and fairness, we often need a way to infer race and ethnicity from names. One way to infer race and ethnicity from names is by relying on the Census Bureau's list of popular last names. The list, however, suffers from at least three limitations: 1. it only contains last names…
We consider training machine learning models that are fair in the sense that their performance is invariant under certain sensitive perturbations to the inputs. For example, the performance of a resume screening system should be invariant under changes to the gender and/or ethnicity of the applicant. We formalize this …
The study tests and optimizes fairness in credit scoring models.
problem Discrimination in credit scoring models based on protected attributes.
method Formal testing and variable identification to optimize fairness and performance.
result Guidance on monitoring and improving algorithmic fairness in credit scoring.
Two simple methods learn fair metrics from data to improve fairness in ML tasks.
problem Lack of widely accepted fair metrics for many ML tasks hinders individual fairness adoption.
method Presented two simple ways to learn fair metrics from various data types.
result Fair training with learned metrics improves fairness on three ML tasks.
The Infant Mortality Rate (IMR) is the number of infants per 1000 that do not survive until their first birthday. It is an important metric providing information about infant health but it also measures the society's general health status. Despite the high level of prosperity in the U.S.A., the country's IMR is higher …
A machine learning model may exhibit discrimination when used to make decisions involving people. One potential cause for such outcomes is that the model uses a statistical proxy for a protected demographic attribute. In this paper we formulate a definition of proxy use for the setting of linear regression and present …
Research shows bias in machine learning can be due to algorithmic flaws, not just data.
problem Underestimation bias in machine learning algorithms.
method Initial research to understand factors contributing to bias in classification algorithms.
result Regularization methods to address overfitting can also accentuate bias.
Financial LLMs need explicit bias consideration to avoid invalid results.
problem Finance-specific biases inflate performance and contaminate backtests.
method Identified five recurring biases and proposed a Structural Validity Framework.
result Explicit bias consideration is necessary for valid deployment claims.
Complex statistical machine learning models are increasingly being used or considered for use in high-stakes decision-making pipelines in domains such as financial services, health care, criminal justice and human services. These models are often investigated as possible improvements over more classical tools such as r…
New model tackles real-world distribution mismatches in machine learning.
problem Real-world applications often have training and test distributions that differ.
method Developed a learning model based on information theory using importance sampling.
result The model performs better under large distribution deviations.
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…
It has been noticed that some external CVIs exhibit a preferential bias towards a larger or smaller number of clusters which is monotonic (directly or inversely) in the number of clusters in candidate partitions. This type of bias is caused by the functional form of the CVI model. For example, the popular Rand index (R…
Mitigates gender bias amplification in model predictions.
problem Gender bias amplification in model predictions.
method Posterior regularization to mitigate bias.
result Almost removes gender bias amplification in model predictions.
Depth uncertainty networks don't improve with bias correction, contrary to expectations.
problem Improving performance in active learning with overparameterised models like NNs.
method Depth uncertainty networks, compared to underparameterised models, show no improvement in performance with bias correction.
result Depth uncertainty networks do not improve with bias correction, unlike underparameterised models.
We quantify causal bias in continuous treatment settings.
problem Identifying and quantifying causal bias in continuous treatment scenarios.
method Developed a novel characterization of causal bias in structural causal models, proving conditions for zero bias and efficient estimation.
result Causal bias can be estimated efficiently under certain structural equation restrictions, allowing for causal regularization of predictive models.
The paper introduces Relative Bias to quantify LLM bias systematically.
problem Quantifying bias in LLMs is challenging due to ambiguity and rapid model emergence.
method Relative Bias framework using Embedding Transformation and LLM-as-a-Judge methodologies.
result The two scoring methods show strong alignment, providing a systematic approach.
Interpolated-MLPs control inductive bias for better performance in low-compute tasks.
problem Low-compute performance gap between MLPs and CNNs.
method Introduced Interpolated MLP (I-MLP) approach to control inductive bias incrementally.
result Continuous logarithmic relationship between inductive bias and performance in low-compute tasks.
SSMs have a built-in bias towards low-frequency components, which can be adjusted.
problem Frequency bias in SSMs affects their performance on long-range sequences.
method Proposed two mechanisms to tune frequency bias: scaling initialization or applying a Sobolev-norm-based filter.
result Tuning frequency bias improves SSMs' performance on long-range sequence learning tasks.
A bias classifier is introduced to resist adversarial attacks.
problem Resisting adversarial attacks on deep neural networks (DNNs).
method Introducing the bias part of a DNN with Relu as the activation function as a classifier, and adding a random first-degree part to make it information-theoretically safe.
result The bias classifier is more robust than DNNs of similar size against adversarial attacks.
UBM transfers bias mitigation from upstream to downstream tasks efficiently.
problem Bias in fine-tuned language models across various tasks.
method Apply bias mitigation to an upstream model, then fine-tune a downstream model on this mitigated model.
result UBM effects transfer to new downstream tasks, creating less biased models.
Active learning introduces bias; this paper fixes it.
problem Bias in active learning due to non-representative training data.
method Formalized bias, identified situations where it's harmful/helpful, introduced corrective weights.
result Corrective weights can improve active learning, especially with overparameterized models.
Bias correction improves language model training performance.
problem Stochastic update bias in preconditioned optimizers.
method Cross-fitted preconditioning and variance-corrected inversion.
result Reduces held-out pretraining loss by 0.15 nats.