Machine Learning with Python: from Linear Models to Deep Learning

6.86x is MITx's machine learning course, part of the MicroMasters Program in Statistics and Data Science and offered at a pace and level of rigour comparable to the on-campus class. It covers the principles and algorithms for turning training data into automated predictions: representation, over-fitting, regularisation, generalisation and VC dimension; clustering, classification, recommender problems, probabilistic modelling and reinforcement learning; and online algorithms, support vector machines, and neural networks and deep learning. Students implement and experiment with the algorithms in several Python projects built around practical applications. The course is instructor-paced and MIT estimates 15 weeks; the catalogue lists 10 to 14 hours per week.

Auditing is free and gives access to the course and materials. The certificate track, priced at $300 on the course page, adds graded assignments and exams and an MIT certificate on completion, and the course can count toward the MicroMasters credential if you go on to complete the programme's other courses and proctored exam. Expect to need probability, linear algebra and Python beforehand, since the course is described as running at MIT's on-campus pace.