Bias and unfairness in machine learning models: a systematic literature review

Pagano, Tiago Palma; Loureiro, Rafael Bessa; Araujo, Maira Matos; Lisboa, Fernanda Vitoria Nascimento; Peixoto, Rodrigo Matos; Guimaraes, Guilherme Aragao de Sousa; Santos, Lucas Lisboa dos; Ramos, Cruz, Gustavo Oliveira; Oliveira, Ewerton Lopes Silva de; Cruz, Marco Aurélio Rodrigues da Cunha e; Winkler, Ingrid G.; Nascimento, Erick Giovani Sperandio

doi:10.48550/arxiv.2202.08176

Cited by 6 publications

(3 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Pagano et al categorize bias mitigation techniques into three main categories: preprocessing, in-processing, and post-processing [68]. Pre-processing approaches work towards rebalancing the data.…”

Section: Step-by-step Approach To Algorithm Biasmentioning

confidence: 99%

A Comprehensive Review of AI Techniques for Addressing Algorithmic Bias in Job Hiring

Albaroudi,

Mansouri,

Alameer

2024

View full text Add to dashboard Cite

The study comprehensively reviews artificial intelligence (AI) techniques for addressing algorithmic bias in job hiring. More businesses are using AI in curriculum vitae (CV) screening. While the move improves efficiency in the recruitment process, it is vulnerable to biases, which have adverse effects on organizations and the broader society. This research aims to analyze case studies on AI hiring to demonstrate both successful implementations and instances of bias. It also seeks to evaluate the impact of algorithmic bias and the strategies to mitigate it. The basic design of the study entails undertaking a systematic review of existing literature and research studies that focus on artificial intelligence techniques employed to mitigate bias in hiring. The results demonstrate that the correction of the vector space and data augmentation are effective natural language processing (NLP) and deep learning techniques for mitigating algorithmic bias in hiring. The findings underscore the potential of artificial intelligence techniques in promoting fairness and diversity in the hiring process with the application of artificial intelligence techniques. The study contributes to human resource practice by enhancing hiring algorithms’ fairness. It recommends the need for collaboration between machines and humans to enhance the fairness of the hiring process. The results can help AI developers make algorithmic changes needed to enhance fairness in AI-driven tools. This will enable the development of ethical hiring tools, contributing to fairness in society.

show abstract

Section: Step-by-step Approach To Algorithm Biasmentioning

confidence: 99%

A Comprehensive Review of AI Techniques for Addressing Algorithmic Bias in Job Hiring

Albaroudi,

Mansouri,

Alameer

2024

View full text Add to dashboard Cite

show abstract

“…Although current facial recognition systems achieve high average accuracy, numerous current improvements are directed towards addressing the disproportionate accuracies across different categories, including race [1][2][3][4]. Many attribute the majority of these issues to unbalanced datasets [5][6][7][8].…”

Section: Introductionmentioning

confidence: 99%

Surveying Racial Bias in Facial Recognition: Balancing Datasets and Algorithmic Enhancements

Sumsion,

Torrie,

Lee

et al. 2024

Electronics

View full text Add to dashboard Cite

Facial recognition systems frequently exhibit high accuracies when evaluated on standard test datasets. However, their performance tends to degrade significantly when confronted with more challenging tests, particularly involving specific racial categories. To measure this inconsistency, many have created racially aware datasets to evaluate facial recognition algorithms. This paper analyzes facial recognition datasets, categorizing them as racially balanced or unbalanced while limiting racially balanced datasets to have each race be represented within five percentage points of all other represented races. We investigate methods to address concerns about racial bias due to uneven datasets by using generative adversarial networks and latent diffusion models to balance the data, and we also assess the impact of these techniques. In an effort to mitigate accuracy discrepancies across different racial groups, we investigate a range of network enhancements in facial recognition performance across human races. These improvements encompass architectural improvements, loss functions, training methods, data modifications, and incorporating additional data. Additionally, we discuss the interrelation of racial and gender bias. Lastly, we outline avenues for future research in this domain.

show abstract

“…In other words, the model relies on heuristics specific to the training data but does not generalize well to unseen data. This can be a significant challenge in real-world machine learning applications [5][6][7] . In the realm of histopathology, distinct categories of biases that arise from various sources 8,9 , as summarized in Fig.…”

Section: Introductionmentioning

confidence: 99%

Bias in Histopathology Datasets: A Comprehensive Investigation on Possible Factors

Kheiri,

Rahnamayan,

Makrehchi

et al. 2024

Preprint

View full text Add to dashboard Cite

Deep Neural Networks (DNNs) have demonstrated remarkable capabilities in medical applications, including digital pathology, where they excel at analyzing complex patterns in medical images to assist in accurate disease diagnosis and prognosis. However, concerns have arisen about potential biases in The Cancer Genome Atlas (TCGA) dataset, a comprehensive repository for digitalized histopathology and a validation source for deep models, suggesting that over-optimistic results of model performance may be due to reliance on biased features rather than histological characteristics. Surprisingly, recent studies have confirmed the existence of site-specific bias in the embedded features extracted for cancer-type discrimination, leading to high accuracy in acquisition site classification. This biased behavior motivated us to conduct an in-depth analysis to investigate potential causes behind this unexpected biased ability toward site-specific pattern recognition. The analysis was conducted on two cutting-edge DNN models: KimiaNet, a state-of-the-art DNN trained on TCGA images, and the self-trained EfficientNet. Considering the classification accuracy of acquisition sites as a measure of bias, a series of experiments has been designed to examine the impact of various factors, including patching, data imbalance, and staining.

show abstract

Bias and unfairness in machine learning models: a systematic literature review

Cited by 6 publications

References 35 publications

A Comprehensive Review of AI Techniques for Addressing Algorithmic Bias in Job Hiring

A Comprehensive Review of AI Techniques for Addressing Algorithmic Bias in Job Hiring

Surveying Racial Bias in Facial Recognition: Balancing Datasets and Algorithmic Enhancements

Bias in Histopathology Datasets: A Comprehensive Investigation on Possible Factors

Contact Info

Product

Resources

About