arXiv research

Research shows bias in machine learning can be due to algorithmic flaws, not just data.

This paper examines AI and ML bias and fairness issues.

Automates bias control in reinforcement learning algorithms.

Study data biases to predict algorithmic discrimination, developing a Data Bias Profile.

Study finds gender bias in human evaluators and shows how machine learning can mitigate it.

ARFF reduces spectral bias in SGD-trained neural networks.

Learning algorithms need bias to generalize and perform better than random guessing. We examine the flexibility (expressivity) of biased algorithms. An expressive algorithm can adapt to changing training data, altering its outcome based on changes in its input. We measure expressivity by using an information-theoretic …

Recommendation systems have been integrated into the majority of large online systems to filter and rank information according to user profiles. It thus influences the way users interact with the system and, as a consequence, bias the evaluation of the performance of a recommendation algorithm computed using historical…

Machine learning algorithms can misrepresent training data, study finds.

Paper analyzes \FedAvg's convergence and introduces a new algorithm to reduce bias.

Federated learning can propagate bias from a few parties to all participants.

Paper corrects bias in online learning algorithms with endogenous data.

Improves decision-making by correcting feature selection bias.

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…

Paper analyzes LSA algorithm bias and error bounds with RR extrapolation.

Study improves statistical power for detecting algorithmic bias in educational data.

A new algorithm removes unexpected correlations in biased data for better clustering.

Building on the view of machine learning as search, we demonstrate the necessity of bias in learning, quantifying the role of bias (measured relative to a collection of possible datasets, or more generally, information resources) in increasing the probability of success. For a given degree of bias towards a fixed targe…

Iterative method 'Concent' corrects spectrum bias in covariance matrices.

We evaluate the folk wisdom that algorithmic decision rules trained on data produced by biased human decision-makers necessarily reflect this bias. We consider a setting where training labels are only generated if a biased decision-maker takes a particular action, and so "biased" training data arise due to discriminato…

Deep RL algorithms can overfit to early experiences, leading to poor performance.

New insights into bias mitigation show DRO isn't a complete solution.

Study shows how steepest descent algorithms' geometric margin increases during training.

Recommendation systems have been integrated into the majority of large online systems to filter and rank information according to user profiles. It thus influences the way users interact with the system and, as a consequence, bias the evaluation of the performance of a recommendation algorithm computed using historical…

This research debiases machine unlearning by using counterfactual examples.

Open problem: Establishing bounds for Cayley-table completion to discover discrete algorithmic axioms.

EBQL reduces bias in Q-learning for improved performance.

We propose an approach to reduce the bias of ridge regression and regularization kernel network. When applied to a single data set the new algorithms have comparable learning performance with the original ones. When applied to incremental learning with block wise streaming data the new algorithms are more efficient due…

This paper proposed a bias-compensated normalized maximum correntropy criterion (BCNMCC) algorithm charactered by its low steady-state misalignment for system identification with noisy input in an impulsive output noise environment. The normalized maximum correntropy criterion (NMCC) is derived from a correntropy based…

Study evaluates bias mitigation methods in deep learning, finds they often exploit hidden biases.

Algorithm corrects bias in classification data.

Mirror Langevin Algorithm converges with zero bias.

Survey on why deep learning works despite having more parameters than data.

WeMix improves data augmentation by correcting bias in deep learning.

New algorithm identifies causal relationships from graphs, even with selection bias.

In classification problems, sampling bias between training data and testing data is critical to the ranking performance of classification scores. Such bias can be both unintentionally introduced by data collection and intentionally introduced by the algorithm, such as under-sampling or weighting techniques applied to i…

Proposes a new algorithm to estimate invariant subspaces across multilayer networks.

New algorithm achieves optimal regret in average reward MDPs without prior bias information.

Algorithm bounds causal queries under selection bias.

New method controls bias in unadjusted Hamiltonian Monte Carlo and underdamped Langevin.

ABROCA assesses algorithmic bias, revealing skewed distributions that inflate results.

We study the problem of learning-to-learn: inferring a learning algorithm that works well on tasks sampled from an unknown distribution. As class of algorithms we consider Stochastic Gradient Descent on the true risk regularized by the square euclidean distance to a bias vector. We present an average excess risk bound …

In high dimensional settings, density estimation algorithms rely crucially on their inductive bias. Despite recent empirical success, the inductive bias of deep generative models is not well understood. In this paper we propose a framework to systematically investigate bias and generalization in deep generative models …

Outlier based Robust Principal Component Analysis (RPCA) requires centering of the non-outliers. We show a "bias trick" that automatically centers these non-outliers. Using this bias trick we obtain the first RPCA algorithm that is optimal with respect to centering.

The least squares Monte Carlo (LSM) algorithm proposed by Longstaff and Schwartz (2001) is widely used for pricing Bermudan options. The LSM estimator contains undesirable look-ahead bias, and the conventional technique of avoiding it requires additional simulation paths. We present the leave-one-out LSM (LOOLSM) algor…

Data-driven algorithms play a large role in decision making across a variety of industries. Increasingly, these algorithms are being used to make decisions that have significant ramifications for people's social and economic well-being, e.g. in sentencing, loan approval, and policing. Amid the proliferation of such sys…

New algorithm offers costless model selection in contextual bandits.

Develops NFCF to reduce gender bias in social media recommendation systems.