LFD Book Forum  

Go Back   LFD Book Forum > Book Feedback - Learning From Data > Chapter 5 - Three Learning Principles

Reply
 
Thread Tools Display Modes
  #1  
Old 10-28-2013, 07:02 PM
Sweater Monkey Sweater Monkey is offline
Junior Member
 
Join Date: Sep 2013
Posts: 6
Default Input normalization

I could use some clarification on the 9th homework assignment.

In normalizing the features I would assume that it is with respect to the bounds on the feature's possible values rather than on the data itself?

For example, the minimum pixel intensity is for the digit data is -1 and the maximum is +1 therefore, using average pixel intensity as a feature, the min and max average pixel intensities that are possible are -1 and +1 (if all pixels were white or all pixels were black for a given data point). So in this case I am under the assumption that the feature doesn't need any shift or scale normalization since the min and max values are already within the [-1,1] bound regardless of the data. Is this correct?
Reply With Quote
  #2  
Old 11-01-2013, 11:22 AM
magdon's Avatar
magdon magdon is offline
RPI
 
Join Date: Aug 2009
Location: Troy, NY, USA.
Posts: 592
Default Re: Input Normalization

Feature normalization (as in the example of data snooping in financial trading in chapter 5) is based on the data itself. Similarly in the assignment, the feature normalization is based on the data.

Quote:
Originally Posted by Sweater Monkey View Post
I could use some clarification on the 9th homework assignment.

In normalizing the features I would assume that it is with respect to the bounds on the feature's possible values rather than on the data itself?

For example, the minimum pixel intensity is for the digit data is -1 and the maximum is +1 therefore, using average pixel intensity as a feature, the min and max average pixel intensities that are possible are -1 and +1 (if all pixels were white or all pixels were black for a given data point). So in this case I am under the assumption that the feature doesn't need any shift or scale normalization since the min and max values are already within the [-1,1] bound regardless of the data. Is this correct?
__________________
Have faith in probability
Reply With Quote
  #3  
Old 11-04-2013, 09:44 PM
alanericy alanericy is offline
Junior Member
 
Join Date: Oct 2013
Posts: 5
Default Plotting Separators

Hi Professor, since the regression will result a highly nonlinear boundary how do we plot it in the feature plot?Thanks!
Reply With Quote
  #4  
Old 11-05-2013, 08:41 AM
magdon's Avatar
magdon magdon is offline
RPI
 
Join Date: Aug 2009
Location: Troy, NY, USA.
Posts: 592
Default Re: Plotting Separators

You just need to show the region of your feature space that is classified as +1 versus -1. This can be done by griding your space and plotting a blue dot for +1 and a red dot for -1.

Quote:
Originally Posted by alanericy View Post
Hi Professor, since the regression will result a highly nonlinear boundary how do we plot it in the feature plot?Thanks!
__________________
Have faith in probability
Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 02:15 AM.


Powered by vBulletin® Version 3.8.3
Copyright ©2000 - 2017, 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.