Coursera has the most reputable online training in Machine Learning (from Stanford U, by Andrew Ng), a fantastic Deep Learning specialization (from deeplearning.ai, also by Andrew Ng) and now a practically oriented TensorFlow specialization (also from deeplearning.ai).
But it has very little offering in Reinforcement Learning, where Coursera clearly lags competition, even though it is hard to find quality online courses for a non-ridiculous price elsewhere. David Silver's recorded lectures at UCL are first class, as is Sutton and Barto's fantastic book, but this is not the same as a coursework with lessons, quizzes, exercises, assignments and perhaps a certificate in the end. Competitors offer online courses and nano-degrees, but it is hard to assess the quality of the teachers and material, and the cost is sometimes unreasonable, to say the least.
I really hope that Coursera will soon produce a specialization up to Andrew Ng's standard on RL, as this is currently and very visibly missing from its catalog. If it was taught by Andrew Ng in person, I would sign up immediately, no questions asked, as, I think, would the thousands of others who followed his offerings in ML and DL.