LFD Book Forum

LFD Book Forum (http://book.caltech.edu/bookforum/index.php)
-   Homework 1 (http://book.caltech.edu/bookforum/forumdisplay.php?f=130)
-   -   Verification versus Learning (http://book.caltech.edu/bookforum/showthread.php?t=4427)

yaser 10-01-2013 04:39 PM

Verification versus Learning
A post at another forum:

In the lecture 2 we find that if we try our hypothesis against a single bin then we call it "Verification" rather than learning.But if we verify different hypothesis against different bins then this "Verification" becomes "Learning". I would like to get an explanation on this particular point.Thanks in advance!
Each bin corresponds to testing one hypothesis against the target function. If there is one bin, what we will be doing is testing a single hypothesis and accepting it or rejecting it according to its performance, with no alternative hypothesis in case we reject it. This is why we describe the situation as "verification" as the only positive outcome is that the hypothesis is verified as a good hypothesis if it is accepted through this process.

If we have multiple bins, these correspond to multiple hypotheses and now we are choosing between different hypotheses according to their performance. Think of this as trial and error until you find a hypothesis that works, and in that case we think of the situation as having "learned."

All times are GMT -7. The time now is 07:26 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.