LFD Book Forum

LFD Book Forum (http://book.caltech.edu/bookforum/index.php)
-   General comments on the course (http://book.caltech.edu/bookforum/forumdisplay.php?f=140)
-   -   Masterfully done (http://book.caltech.edu/bookforum/showthread.php?t=1156)

samirbajaj 08-25-2012 11:42 AM

Masterfully done
 
I've enjoyed every lecture in the course so far, but I have to say that the one on kernel methods is very well done. I've tried to read about kernel methods on my own, but lacking the math sophistication, I've never really understood how they worked. I think I do now.

Please consider making more of your courses available to the world. You have a valuable gift.

Thanks.

-Samir

yaser 08-25-2012 11:53 AM

Re: Masterfully done
 
Quote:

Originally Posted by samirbajaj (Post 4428)
I've enjoyed every lecture in the course so far, but I have to say that the one on kernel methods is very well done. I've tried to read about kernel methods on my own, but lacking the math sophistication, I've never really understood how they worked. I think I do now.

Please consider making more of your courses available to the world. You have a valuable gift.

Thanks.

-Samir

Thank you for your kind words.

gah44 09-04-2012 11:58 AM

Re: Masterfully done
 
Thanks for everything!

Having graduated about 30 years ago, not counting graduate school,
it was interesting to have a homework deadline after so many years.

You could have done it without any deadlines, but it wouldn't feel
like a real course!

And thanks to Caltech for supporting it.
(And YouTube for hosting the lectures.)

luis_lfd 09-05-2012 07:24 PM

Re: Masterfully done
 
The same here. I really enjoyed this course and professor Yaser's lectures are fantastic.

Hope to see more on-line classes delivered by such great teachers.

gah44 09-06-2012 12:31 AM

Re: Masterfully done
 
I was just wondering, though, does the real Caltech course have an
unlimited (well, two weeks) time open book final?

I would have expected a three hour, closed book (honor system)
take home final.

Not that I am complaining, but just wondering.

yaser 09-06-2012 08:40 AM

Re: Masterfully done
 
Quote:

Originally Posted by gah44 (Post 4969)
I was just wondering, though, does the real Caltech course have an
unlimited (well, two weeks) time open book final?

I would have expected a three hour, closed book (honor system)
take home final.

Not that I am complaining, but just wondering.

There is no final given at Caltech for this course. Just open-book homeworks from which the multiple choice version was derived whenever possible.

itooam 09-08-2012 05:36 AM

Re: Masterfully done
 
Just wanted to echo the sentiment expressed above. I have found the course extremely useful/helpful and would certainly recommend it to others.

The things I would suggest as a pre-requisite to others considering this course would be a basic understanding of linear algebra and a tool for easy implementation I.e., Octave.
I believe this would make the course accessible to many more people. Had I not done a crash course in the above beforehand I would have given up after perhaps the first lecture. So glad I persisted... esp when the going got tough. So very rewarding. The course got better as time progressed.

I would like to thanks all those involved with making this course free and especially the time the lecturers set aside in answering our
questions on the forums. Direct help/feedback is invaluable.

Many many thanks :)

Willem 09-09-2012 02:23 PM

Re: Masterfully done
 
Thank you professor Abu_Mostafa for a very stimulating course! The lectures were among the best I ever had, thank you for this! I also want to mention the well thought out homeworks, which indeed made us roll up our sleeves! Thank you again, I can only agree with the OP that I hope in the future there will be another course :)

jiunjiunma@gmail.com 09-10-2012 03:40 PM

Re: Masterfully done
 
I concur. The homework is really well thought out and reinforces the learning (no pun intended). I especially like that it doesn't dictate the use of any specific language so you can pick your own tool.

Thank you so much Professor Abu-Mostafa. This is the best online machine learning class and you really give us a map (with tons of insights) to explore further. I will certainly miss the weekly "welcome back"...

yaser 09-10-2012 08:31 PM

Re: Masterfully done
 
You guys are very kind. I appreciate it.

AdrienL 09-12-2012 12:41 AM

Re: Masterfully done
 
I will add my own words to those already told here : thank you Professor and thanks to all the people who made this possible (the TAs, staff, ...).

The lectures were excellent and delivered with exceptional clarity by the lecturer :) — I especially enjoyed the well designed slides, the right balance between concepts, their underlying mathematics and illustrative examples. The final Q&As and the exchanges on the discussion forums with the Professor, the TAs and my fellow students helped clear out any interrogations that I had and provided me with more insight on interrogations I did not have :).

I don't know if you plan to make other re-runs of this course, but I would definitely advise it to anyone interested in the subject.

Perhaps my only grieves would be, that it took me quite a lot of time to learn new packages towards the end of the course (quadratic programming and libsvm, whose documentations were poor — hopefully the posts on the discussion forums helped me tremendously in dealing with these libraries), and that the interface summarizing the answers on the answer page is not very clear : I usually have to write down my answers on a sheet of paper and the correct answers from the answer key to figure on which I was wrong and on which I was right.

But these are only minor grievances, so, once again, thanks for providing us with this high quality free online course on machine learning.

(I still have to complete the final though ! :) But after that I will personally try to follow coursera's upcoming probabilistic graphical models course, as well as the course on neural networks and a few others, if my free time allows it :))

janjetin 09-13-2012 09:21 AM

Re: Masterfully done
 
I'd like to add my own thanks and commendations to Prof. Yaser Abu-Mostafa and others who are responsible for this course.

The lectures were particularly well conducted and are probably the best lectures I've watched so far, compared to other online courses. Prof. Abu-Mostafa explains the material in clear and very interesting manner, emphasizing the important points that are useful in practical implementations.

Homework and final problem sets were challenging, and enabled me to learn from my mistakes and use the correct answers to correct my hypothesis.

The only thing missing from the course is a large scale programming project dealing with a real learning situation, (i.e. something like the Netflix competition related collaboration the professor talked about in the last lecture).

I'm looking forward to seeing and attending more courses from prof. Abu-Mostafa and Caltech in the future.

Alejandro Martinez 09-14-2012 01:06 PM

Re: Masterfully done
 
Thanks prof. Mostafa!

I really liked the course and the book, both very well done. I think the balance between foundations and practical aspects is really good.

If there are plans for a second part, I would like to participate.

Thanks again, Alejandro.

ramin 09-15-2012 09:57 AM

Re: Masterfully done
 
What a gem this class is! Thanks to the professor for presenting the concepts jargon-free.

fgpancorbo 09-17-2012 05:28 PM

Re: Masterfully done
 
I second the thoughts of the others here. This was a great class and I do believe that prof Abu-Mostafa is one of the best instructors around! My sincere thanks to him and the rest of the staff at Caltech, NTU and RPI for making this experience possible!!!

kdresser 09-18-2012 04:39 AM

Re: Masterfully done
 
+1 to all the above.

I learned more here, both foundational and practical, than in any BSc course I took many decades ago. This course was challenging and rewarding, especially the homeworks. Whew.

I thank Prof. Yaser Abu-Mostafa for creating and teaching this course in his inimitable skill & style, Prof. Malik Mogdan-Ismail and Prof. Hsuan-Tien Lin for their contributions to the discussion threads, all the TA's, and Caltech & its supporters for this valuable course.

Definitely masterfully done.

JKD


All times are GMT -7. The time now is 09:01 PM.

Powered by vBulletin® Version 3.8.3
Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
The contents of this forum are to be used ONLY by readers of the Learning From Data book by Yaser S. Abu-Mostafa, Malik Magdon-Ismail, and Hsuan-Tien Lin, and participants in the Learning From Data MOOC by Yaser S. Abu-Mostafa. No part of these contents is to be communicated or made accessible to ANY other person or entity.