View Single Post
  #1  
Old 04-17-2012, 02:39 AM
rohanag rohanag is offline
Invited Guest
 
Join Date: Apr 2012
Posts: 94
Default question target distribution

In the lectures, we take a deterministic target function and add noise to it to make a target distribution. This is what we have done in the programming assignment too. But is it advisable to introduce noise in a data set which is handed to you? Maybe the data set handed over to you already has noise, maybe not?

One more question, in the lecture, we said noise = y - f(x) ,
Noisy target = deterministic target f(x) = E(y|x) + Noise (y - f(x) )

I do not get the above equation, shouldn't
Noisy target = deterministic target + Noise
Noisy target = f(x) + Noise ( y - f(x) )
Noisy target = y

This is really confusing me.
Reply With Quote