Looking beyond accuracy to improve trust in machine learning
01/09/18
Because complex machine learning models are essentially black boxes and too complicated to understand, we need to use approximations to get a better understanding of how they work, e.g. with LIME. Read more