Well, I haven't taken all of them, so I can't give you a direct comparison. But I am pretty sure Andrew Ng's Coursera one still holds up. I'm doing the Geoffrey Hinton one on Neural Networks now and it's been good. I've also been working through the Udacity ML course and there's some good stuff in there (lots of sci-kit learn stuff, so if you're especially interested in that, it's definitely good).
And not "Machine Learning" per-se, but related, there's a series on Coursera, from Duke, on "Probability and Statistics with R" that I've been going through. I'm on the Bayesian Statistics module now and I've learnt a ton from this sequence.
And not "Machine Learning" per-se, but related, there's a series on Coursera, from Duke, on "Probability and Statistics with R" that I've been going through. I'm on the Bayesian Statistics module now and I've learnt a ton from this sequence.