View Single Post
  #4  
Old 07-18-2012, 09:12 AM
yaser's Avatar
yaser yaser is offline
Caltech
 
Join Date: Aug 2009
Location: Pasadena, California, USA
Posts: 1,478
Default Re: Hoeffding’s inequality

Quote:
Originally Posted by itooam View Post
from the lectures I understood Eout(g) to be the "true" error (if you had access to the full population of data). In reality, when we are applying a machine learning technique, we do so because we have limited data and therefore I fail to understand how this formula can be applied?
In a practical machine learning situation, we don't have E_{\rm out} but we estimate it using a test set (set of fresh points) as we have done in Homework 1 with the perceptron. The accuracy of that estimate will be discussed in detail when we talk about validation.
__________________
Where everyone thinks alike, no one thinks very much
Reply With Quote