If you plan to take ATML before or without taking ML-A and ML-B, please, read this carefully and check our self-preparation instructions.

In general it is advised to take ML-A and ML-B before taking ATML. ML-A provides a general overview of machine learning and ML-B provides a more theoretical overview of machine learning, whereas ATML goes in depth, further, into a number of selected topics including advanced analysis of learning algorithms and differentially private algorithms. However, if you have a strong mathematical background and already familiar with the field, you can take ATML without or before taking ML-A and ML-B. Typically students from statistics, math-economics, actuarial math and similar fields have no problems taking ATML before or without taking ML-A and ML-B. If your math skills are not on par with students from stats or math and proving theorems is not your hobby, it is strongly advised to take ML-A and ML-B first. 

Please, be aware that our ML courses (both ML-A and ML-B) are more theoretical than average machine learning courses. So if you have taken a machine learning course elsewhere it does not necessarily prepare you for taking ATML. Solve the exercises under self-preparation instructions to see if you are fit to take ATML.