LFD Book Forum

LFD Book Forum (http://book.caltech.edu/bookforum/index.php)
-   Chapter 1 - The Learning Problem (http://book.caltech.edu/bookforum/forumdisplay.php?f=108)
-   -   Solutions to exercises (http://book.caltech.edu/bookforum/showthread.php?t=4882)

napoca-ai 04-01-2019 07:08 PM

Solutions to exercises
 
Hello,

The course and book are great! I've been going through a number of ML courses and books but this is definitely the best. Thank you Prof Yaser!

Can I find somewhere the solutions for the exercises in the book? It'd be super useful to check my answers against them as I go through the book.

Thanks again for all this, Prof Yaser!

LR

htlin 04-04-2019 02:02 AM

Re: Solutions to exercises
 
There is no solution manual pdf for distribution. But you are very welcomed to discuss about your answers with the community here.

ppaquay 04-17-2019 05:16 AM

Re: Solutions to exercises
 
Hi,

You could also check out my solutions to the book problems on GitHub :

https://github.com/ppaquay/Learning-from-Data-Solutions

Best regards !

Roelof 02-23-2021 06:35 AM

Re: Solutions to exercises
 
In problem 1.1 the bags and the suggested use of Bayes Theorem is a bit of a red Herring. You have 4 balls 3 of which are black. You pick one ball which is black. There are now 3 balls 2 of which are black. The chance that you pick a black one is 2 out of 3. It doesn't matter how many bags there are (as long as there is at least one ball in each bag) or from which bag you choose the second ball (as long as there is a ball in it).

So for example if I have 10 balls spread over 3 bags (e.g. 4,4,2) and one is white and the first ball I pick is black. Then the chance I pick another black ball from any particular bag is 8/9. Try solving this using the Bayesian Theorem :^).

You can think of the bags as groupings on the balls on a table in the dark. If you select a ball from one group it doesn't matter which group you select the second ball from.

Roelof 02-23-2021 08:09 AM

Re: Solutions to exercises
 
Quote:

Originally Posted by ppaquay (Post 13256)
Hi,

You could also check out my solutions to the book problems on GitHub :

https://github.com/ppaquay/Learning-from-Data-Solutions

Best regards !

In problem 1.3(c)
you make the step:

||w(t-1) + y(t-1)x(t-1)||^2 = ||w(t-1)||^2 + ||x(t-1)||^2 +2 y(t-1)w^T(t-1)x(t-1)

What lemma are you basing this step on?

You can prove by simple algebraic manipulation (so no need for use of a lemma) by using w(t)=w(t-1)+y(t-1) and the definition of ||w||^2 namely, w^T w


All times are GMT -7. The time now is 01:54 AM.

Powered by vBulletin® Version 3.8.3
Copyright ©2000 - 2021, 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.