How to Prevent Bias in Machine Learning – Becoming Human: Artificial Intelligence Magazine

The following article is based on work done for my graduate thesis titled: Ethics and Bias in Machine Learning: A Technical Study of What Makes Us "Good," covering the limitations of machine learning algorithms when it comes to inclusivity and fairness. As Cathy O'Neil discusses in her book, Weapons of Math Destruction, the seeming impenetrability and absolute value of machine learning may not be all that we bargained for. Though machine learning appears to indisputably increase business value and efficiency, in some cases, it can sow inequality deeper by hard-coding it into our machines. It is imperative that machine learning experts, creators, and contributors account for "doing the right thing" as much as they do "meeting the bottom line" to balance the enormous power these mechanical decision makers possess. A machine learning algorithm is typically code written by a data scientist in a programming language such as R, Python, or Javascript.