View Single Post
Old 06-10-2013, 04:37 PM
alasdairj alasdairj is offline
Join Date: Mar 2013
Posts: 12
Default Top Ten Reasons I Loved This Course

10. I really liked the way the assignments did not have "starter code", meaning that you really understood the solution start to finish as you wrote the whole thing yourself. Other MOOCs I have taken make you fill in the blanks which is a lot less satisfying - you learn a lot more by writing it yourself, I think.

9. Related to the above, I also really liked the way we generated data ourselves and repeated experiments for many of the questions. This gives you a better feel for what the "real world" data is doing - namely sampling from the target function, f, using some probability distribution.

8. I liked having a companion book which was very aligned to the course content, so I could read up on topics as needed - and having additional problems to dig deeper as needed.

7. Having the class ranking given for each assignment along with the cumulative ranking as you progress gave a huge incentive to "do well".

6. The community in the forums has been one of the best - people gave very constructive help in understanding the material.

5. The lecture formats - I much preferred the "recorded before a live studio audience" approach here over the usual MOOC "studio recorded" lectures. I'm not sure why, but it seems to make the lectures easier to follow.

4. The approach of the course in attacking the real fundamentals of the subject was very helpful to me - somewhat like reading Knuth's Art of Computer Programming books or Spivak's calculus text. The other ML MOOCs I took built on these foundations without explaining them.

3. Having Professor Abu-Mostafa contribute to the forum discussions was an extremely surprising and welcome bonus to this course. No other MOOC I have taken has had direct involvement by the instructor in the forums. Thank you, Yaser!

2. The material was exceptionally well taught! Somehow Yaser took you through the steps you needed in order to grasp the concepts and the detail. I can only compare this to the writings of Feynman - both of these Caltech professors give you the confidence that you, too, can understand the material - and follow through with their promise - I still don't know how!

1. This course has inspired me to start a machine learning project, and to look for career opportunities in that direction!
Reply With Quote