View Single Post
  #2  
Old 10-18-2013, 12:29 AM
yaser's Avatar
yaser yaser is offline
Caltech
 
Join Date: Aug 2009
Location: Pasadena, California, USA
Posts: 1,478
Default Re: [Lec-3] Include w0 and x0 in linear regression formula

Quote:
Originally Posted by blunderboy View Post
Take the credit card example again..By Linear Regression output, I want to know how much credit amount should be approved for a particular customer.

On Slide 10/23,
Linear Regression output h(x) = Summation(i=0 to d) wi xi = wx

This means we are including x0 and w0 in linear regression output. I am not able to understand why do we need to include 0th data where x0 = 1.
This is to allow the linear formula to have a bias term. If the optimal solution does not require a bias term, the algorithm will set w_0 to zero, so there will be no harm done.
__________________
Where everyone thinks alike, no one thinks very much
Reply With Quote