Before a machine learning model is deployed, its performance is evaluated. However, apart from the performance aspect, it is also important to know what the model has learnt. This is necessary so as to ensure that the model hasn’t learnt something discriminatory or biased. [Related Article:... Read more