Algorithmic Factors Influencing Bias in Machine Learning
Not on the map. Why?
Abstract (arXiv)
It is fair to say that many of the prominent examples of bias in Machine Learning (ML) arise from bias that is there in the training data. In fact, some would argue that supervised ML algorithms cannot be biased, they reflect the data on which they are trained. In this paper we demonstrate how ML algorithms can misrepresent the training data through underestimation. We show how irreducible error, regularization and feature and class imbalance can contribute to this underestimation. The paper concludes with a demonstration of how the careful management of synthetic counterfactuals can ameliorate the impact of this underestimation bias.
Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 8 of its 25 references that are in the map. First posted on arXiv 28 Apr 2021.
Builds on
8 of 25 references.
References · 8
- 2016Equality of Opportunity in Supervised Learningcited by 2,000
- 2017Counterfactual Fairnesscited by 802
- 2014Certifying and removing disparate impactcited by 715
- 2019What Do Compressed Deep Neural Networks Forget?cited by 164
- 2020Characterising Bias in Compressed Modelscited by 162
- 2020Good Counterfactuals and Where to Find Them: A Case-Based Technique for Generating Counterfactuals for Explainable AI (XAI)cited by 79
- 2016Auditing Black-box Models for Indirect Influencecited by 41
- 2020Underestimation Bias and Underfitting in Machine Learningcited by 2
Built on by
Citing papers per year: 2021: 2; 2022: 3; 2023: 1; 2024: 3.
9 papers.
- 2022Understanding Unfairness in Fraud Detection through Model and Data Bias Interactionscited by 7
- 2022Stock Embeddings: Learning Distributed Representations for Financial Assetscited by 4
- 2022Prisoners of Their Own Devices: How Models Induce Data Bias in Performative Predictioncited by 4
- 2021Using Pareto Simulated Annealing to Address Algorithmic Bias in Machine Learningcited by 3
- 2024Trustworthy and Responsible AI for Human-Centric Autonomous Decision-Making Systemscited by 3
- 2024Enhancing Fairness and Performance in Machine Learning Models: A Multi-Task Learning Approach with Monte-Carlo Dropout and Pareto Optimalitycited by 2
- 2024Outlier Detection Bias Busted: Understanding Sources of Algorithmic Bias through Data-centric Factorscited by 2
- 2021Introducing a Family of Synthetic Datasets for Research on Bias in Machine Learningcited by 1
- 2023A Comparative Analysis of Bias Amplification in Graph Neural Network Approaches for Recommender Systemscited by 0
- 2024Trustworthy and Responsible AI for Human-Centric Autonomous Decision-Making Systems28 Aug 2024
- 2024Outlier Detection Bias Busted: Understanding Sources of Algorithmic Bias through Data-centric Factors24 Aug 2024
- 2024Enhancing Fairness and Performance in Machine Learning Models: A Multi-Task Learning Approach with Monte-Carlo Dropout and Pareto Optimality12 Apr 2024
- 2023A Comparative Analysis of Bias Amplification in Graph Neural Network Approaches for Recommender Systems18 Jan 2023
- 2022Understanding Unfairness in Fraud Detection through Model and Data Bias Interactions13 Jul 2022
- 2022Prisoners of Their Own Devices: How Models Induce Data Bias in Performative Prediction27 Jun 2022
- 2022Stock Embeddings: Learning Distributed Representations for Financial Assets14 Feb 2022
- 2021Introducing a Family of Synthetic Datasets for Research on Bias in Machine Learning19 Jul 2021
- 2021Using Pareto Simulated Annealing to Address Algorithmic Bias in Machine Learning31 May 2021