FAIRNES AND BIAS IN MACHINE LEARNING MODELS

In recent decades the volume of data generated by businesses and consumers has rocketed, from information on location, buying habits, browsing activity and more. With this data boom comes the opportunity to exploit that data for commercial gain. Machine learning is the way to do this, an actively developing field, with improvements to speed and scalability happening at pace. With these come the risks of biases in the data or the models used to exploit them. As with all advancements, the understanding of these risks is still developing, and care must be taken to both measure and mitigate them. © 2023 Institute of Telecommunications Professionals. All rights reserved.