I took Geoff Hinton's Coursera course about 3 years ago, and it remains my favourite online course I've ever taken. While I do understand that it may have felt a bit theoretical and heavy, I also think that this theory is more important to understand in neural networks than in other areas of ML e.g. for debugging issues. I also loved his wry sense of humour and turn of phrase.
I felt that Andrew Ng left out a lot of the details that I was curious about. For example, I remember he introduces logistic regression and cross-entropy, but he kind of just writes down the equations and says they differentiate nicely, he never explains where they actually come from, or even points the interested viewer where to look for more information (short answer : information theory, also this post https://terrytao.wordpress.com/2016/06/01/how-to-assign-part... is a great explanation for cross entropy)
> While I do understand that it may have felt a bit theoretical and heavy, I also think that this theory is more important to understand in neural networks than in other areas of ML e.g. for debugging issues.
This is the same argument people gave for Machine Learning but Andrew Ng showed us otherwise.
> I felt that Andrew Ng left out a lot of the details that I was curious about.
This is exactly the beauty of Ng's ML course. The fact that Andrew Ng made you curious makes CS229A (Applications of Machine Learning) a very successful course. Now, if you want to satiate that curiosity I will recommend you CS229 [0]. It contains all the theory and math and YET that course is as beautiful as CS229A. This further proves the point that you do not need theory to be boring to achieve your objectives. If you want to teach something, you will find a way. Au contraire, Jeff Hinton's course makes me not want to even touch or understand Neural Networks again. FYI, I have a Ph.D in Computer Science and I work as a Data Scientist. If this is my view, imagine what it would be like beyond the us elites?
> FYI, I have a Ph.D in Computer Science and I work as a Data Scientist. If this is my view, imagine what it would be like beyond the us elites?
The probability, linear algebra and calculus material required to understand this should be covered in undergraduate courses. Which part exactly did you find so hard?
This isn't about the content but the structure and explaining the intuition behind algorithms. There are insights behind every algorithm which needs to be put down explicitly in simple English; not mathematical symbols. A course should be accessible to everyone given the prerequisites. Just because a course is easy for me, does not mean its a good course. The Hinton course is not palatable and you can clearly see that in the discussion forums.
Andrew Ng's CS229 course goes beyond the mathematics and explains the magic behind these algorithms. Unfortunately, I know many people who use the mathematics trope to not share these insights and keep them closely guarded to their chest.
I remember taking Ng's course soon after having taken a similar course. While his offered insights were occasionally useful, I found a number of them distracting, especially when they differed substantially from what I had learned in the other course. I found most of his explanations simple but some of them were unsatisfactory, as the suggested motivations did not imply the taught solutions. This was forgivable where I knew of several alternatives, but not when the information was new to me.
His class left me thinking I was able to apply problems to solutions, not the other way around. I will agree that he was a better teacher than Norvig and Thrun, and more engaging (though less holistic) than Widom.
I felt that Andrew Ng left out a lot of the details that I was curious about. For example, I remember he introduces logistic regression and cross-entropy, but he kind of just writes down the equations and says they differentiate nicely, he never explains where they actually come from, or even points the interested viewer where to look for more information (short answer : information theory, also this post https://terrytao.wordpress.com/2016/06/01/how-to-assign-part... is a great explanation for cross entropy)