Machine learning allows computers to learn a model for a given task, such as face recognition, with a high degree of accuracy, using data. However, after these models are generated, they are often treated as black boxes by developers and the limitations of a model are often unknown to end-users.
