Data Bias and Its Impact on Machine Learning

Data-driven algorithms serve as the central power in many applications that have multifaceted effects on people’s lives. These algorithms learn patterns from the input data, and they inevitably inherit and amplify issues with the data. Many of these issues trace back to inequality in our society. Dr. Ke Yang, a postdoc fellow at the University of Massachusetts Amherst, will join MIT Horizon to discuss the types of data bias that cause machine-learning models to produce unfair and untrustworthy predictions and techniques to mitigate those unfair predictions.

Register on Zoom: https://mit.zoom.us/webinar/register/WN_pTVQ-X85QJudfF02IS_KZQ