Lecture 17 - Three Learning Principles

[@AlexanderPolomodov]

Именно та лекция в курсе, которую стоит посмотреть, даже если на весь курс не хватило времени:) Превосходная лекция, превосходный лектор.

[@HajjAyman]

Your enjoyment of the subject matter shows on your face! I can't blame you; it's a very fascinating subject. Thanks for making it easy to understand.

[@QingkaiKong]

Thanks for the great lectures. I like this saying 'if you torture the data long enough, it will confess'

[@brainstormingsharing1309]

Absolutely well done and definitely keep it up!!! 👍👍👍👍👍

[@MrCmon113]

44:30 I'm not bad with money, I'm a support vector for credit approval.

[@mudafaaedilecekbırseyyok]

hahaahahahahaha

[@kritikseth]

around a million students started this course, but only 70k students finished it.

[@acgalt]

The Occam's razor trimmed them away...

[@AndyLee-xq8wq]

great! tho I'm still not very clear with reuse of dataset

[@rachelshitrit]

hi, did somebody understand why the is the growth function in the postal cram equal to 1?

[@ksawery6568]

To us it appears as though we have the perfect hypothesis. There is only one possible dichotomy for any N samples (extremely simple model), and its the one that predicts the results of the game. In actuality, the hypothesis is completely random, so it could generate any dichotomy just by chance. So actually the growth function is equal to the number of possible dichotomies, 2^N.

[@josejoelmartinezcardoza8926]

Son pocos mis comentarios porque no se inglish pero mientras sea tecnología megusta

[@brainstormingsharing1309]

👍👍👍👍👍👍👍

[@pascal45678]

what does VC stand for?

[@laurin1510]

vapnik chervonenkins, the guy that extended höfdings inequality for the general case with multiple hypothesis.

[@MrCmon113]

@@laurin1510 Vapnik and Chervonenkins are two people. 😂

[@movax20h]

8:55, about Occam's Razor. It is not quite correct. The Occam's Razor principle, is that the model with smallest number of assumptions about data and real behavior of unseen data is most plausible. "Entities must not be multiplied beyond necessity" (Non sunt multiplicanda entia sine necessitate). Or Numquam ponenda est pluralitas sine necessitate [Plurality must never be posited without necessity]. The princple is much older than the Occam, but he was the one to popularise it in the west in more formal way. That doesn't mean that the model must be the simplest. If the model is more complex, but have less assumption, and it fits the data, it is better.

[@Bing.W]

Here "simpler model" refers to "less assumptions", hence no need to repeat the tautology.

[@icstdss2139]

Hay...