Abstract
Over the last decade, the importance of machine learning increased dramatically in business and marketing. However, when machine learning is used for decision-making, bias rooted in unrepresentative datasets, inadequate models, weak algorithm designs, or human stereotypes can lead to low performance and unfair decisions, resulting in financial, social, and reputational losses. This paper offers a systematic, interdisciplinary literature review of machine learning biases as well as methods to avoid and mitigate these biases. We identified eight distinct machine learning biases, summarized these biases in the cross-industry standard process for data mining to account for all phases of machine learning projects, and outline twenty-four mitigation methods. We further contextualize these biases in a real-world case study and illustrate adequate mitigation strategies. These insights synthesize the literature on machine learning biases in a concise manner and point to the importance of human judgment for machine learning algorithms.
Highlights
Over the last decade, insights obtained from machine learning (ML) embedded in artificial intelligence (AI) revolutionized and fundamen tally changed almost every aspect of daily life.1 For example, ML algo rithms make movie recommendations, suggest products to buy, decide on loan applications, and influence hiring decisions (Bogen & Rieke, 2018; Cohen et al, 2019)
The outcome of the ML model influences the training data such that a small bias can be reinforced by a feedback loop
A feedback bias arises when the outcome of the ML model influences the training data such that a small bias can be reinforced by a feedback loop
Summary
Insights obtained from machine learning (ML) embedded in artificial intelligence (AI) revolutionized and fundamen tally changed almost every aspect of daily life. For example, ML algo rithms make movie recommendations, suggest products to buy, decide on loan applications, and influence hiring decisions (Bogen & Rieke, 2018; Cohen et al, 2019). While some literature reviews of ML and AI in business and marketing already exist (e.g., Guha et al, 2021; Huang & Rust, 2021; Puntoni et al, 2021), all these studies only briefly touch on ML biases. ML biases should be presented in a comprehensible manner so that marketing researchers and practitioners can effectively manage and address them in their ML projects. We address these gaps by reviewing the largely disconnected liter ature on ML biases from different fields and by providing a shared ter minology of mitigation methods to prevent these biases. Thereby, we hope to inform and sensitize re searchers and managers alike about the specific biases that might be present in their ML projects and the methods to mitigate their impact
Sign-in/Register to view highlights & summary 
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call