Wow, what a ride!

[alfansome]

Just wanted to thank everyone involved, especially Prof. Abu-Mostafa for a first-rate course. Having not been a student for many,many years its amazing to see the advancements in problem-solving technology that are more or less taken for granted for today's students. If I'd had to do (most of) these problems using the standard-issue slide rulers that I had as an undergraduate, it certainly would have taken much more time Even though you need to get the theoretical stuff down, its so much easier to understand some of the concepts when you can do the computational modeling thats available on your home computer or laptop (or even iPad) with very little effort (after the initial learning curve for the math packages).

[mathprof]

I am very grateful for the opportunity to have taken this class, and wish to thank our prof Yaser Abu-Mostafa and the TAs and anyone else involved in making this course available. This has been a wonderful learning experience. And to be able to take this sitting in the comfort of my own living room.. How cool is that?!


Last fall semester I took Andrew Ng's ML class (up until this class I would have said that it was the best course I ever took; now it's a toss-up!). This class has very nicely complemented what prof Ng covered there. It was nice to see more of the mathematics behind the algorithms. Also I've enjoyed the HW (once I got over feeling I needed to get 100% on everything), and found those problems to be an excellent learning experience. The insights Yaser has shared have been both profound and will prove profoundly useful in applying ML to real-world data. I look forward to pursuing further studies in ML. What fun!

I'm sure that many are curious about the future status of this class... Will this course be offered again in the future? Are there plans to offer any other Caltech classes online like this?

[shockwavephysics]

I concur. This opportunity was amazing, and I thank Professor Abu-Mostafa very much for making this course available online. The homework took a substantial amount of time, but was well worth it. I am sad that we did not build a Bayesian classifier, but there is nothing stopping me from doing so on my own. I also want to thank all the people who have been posting on this forum. It is a wonderful way to connect with other people and talk about the coarse material. I wish we had been able to do that when we were students. Of course, we got together in person with friends, but I feel this helps create a much larger group of friends, and their collective experiences do help a larger fraction of people spend less time going down incorrect paths. Also, it has been fun!

[kkkkk]

I would like to thank the profs, staff and donors who have made this course a very educational adventure. This course gives me a good starting foothold in a very steep topic.

Since this is an online class, I have found the forum to be very helpful especially with the homework. I appreciate the time some posters took to answer my queries, e.g. regarding VC dim and growth function. My personal best post award goes to the guy/gal who posted the possibly right answer to hw4 qn 10.

I now look forward to a break from spending my last 8 weekends on the homeworks

[vpoenisc]

A little belatedly I want to join the large group of people thanking Caltech for making this class publically available and Yaser for his fantastic job in teaching it.

I'm in industry, hoping to apply what I have learned on a couple of upcoming projects this year. I found the course content extremely worthwhile.

I find Machine Learning difficult, because it is counter-intuitive to me. Even after the class, I find it hard to rid myself of the notion of applying all possible cleverness to somehow explain the available data and then to extrapolate that explanation, i.e., as we learned, the notion to minimize the in-sample error without regard to the out-of-sample error. And the Netflix example, to predict people's preferences for movies without any input of domain expertise on movies or psychology, that's still black magic, even after understanding how it's done.

While it's easy to pick up a couple of algorithms from the many text-books and online materials out there, it is the solid foundation, both mathematical and practical, as well as this better intuition that I would have missed studying alone without this class. Also, the cadence of the lectures and the homeworks helped to enforce some learning discipline.

Regarding the details of the course content and their overlap with my planned projects, I would have liked some more discussion of other practical problems. One is methods to handle classification problems where one result occurs much more frequently than the other. Another is methods for data sets that are not well described by tabular x and y's, but more naturally through a graph structure.

On the other hand, I cannot really think of any topic I would have liked to have been dropped from the course. And I imagine that the unhurried pace of the lectures was one of the magic ingredients that made the course so successful.

In summary, not many suggestions how to improve a class of such high caliber. Thanks again.

[cambridge11]

I thoroughly enjoyed the class!

I was wondering if we can receive any kind of certificate or transcript for the course? I know Udacity provides a certificate on the completion of a course.

Thanks!

[Hillbilly]

Originally Posted by cambridge11

I was wondering if we can receive any kind of certificate or transcript for the course? I know Udacity provides a certificate on the completion of a course.

As you probably know already, Udacity is a business, that offers courses on its own, independently of any affiliation with universities. So it can do whatever it wants. Free online courses like Learning From Data, with an existing university involved, raise many complex issues when it comes to issuing certificates. This was a big controversy at Stanford last year, when some profs there started up free online courses. Those profs went on to found Udacity, and a competing business Coursera, each with different philosophies on the certificate question. The MITx initiative is yet another approach.

I agree it would be great if the LFD staff could issue us a certificate, or something, or even an email, acknowledging our completion. A bit of thought reveals why it's not easy. What would such a document mean, exactly? What does it allow the student to legitimately claim? What kind of misunderstandings and misrepresentations might occur? What's to prevent just anyone from faking the document? Would it dilute the stature of genuine credits and degrees from Caltech? Will Caltech be inundated with requests from Human Resources departments requesting verification of the document? And so on.

Caltech and Professor Abu-Mostafa are pioneering something new with LFD. Maybe all those issues above will get worked out, eventually. I'm happy just to have had the experience.

[StevenLau]

I totally agree with Alfansome ... learning in these days are so advanced compared with that in our old days ... awesome ...

[jianpan]

I enjoyed this class so much that I did the last three lectures and homework on my vacation to Asia this summer. Now this class will always be in the same memory as my trip. Thank you Prof. Abu-Mostafa for a fantastic learing opportunity!