02-14-2016 01:05 AM
In linear regression we have y=b0 +b1x
where bo is the itercept and b1 is the slope.
b1=y2-y1/x2-x1 ,by this we get the slope
and b0 we get by putting the x=0
In multiple linear regression we have Y=b0+b1x1+b2x2+b3x3
now here how can we get the intercept and slope??
For intercept do we need to put x1,x2 and x3 all to 0??
02-14-2016 04:51 AM - edited 02-14-2016 04:51 AM
The coefficients of each x term are the slopes.
If you set all the x to zero, you can calculate the intercept.
This explains it well: https://www.st-andrews.ac.uk/media/capod/students/mathssupport/Simple-linear-regression.pdf
02-14-2016 08:46 PM
No. You can't . It is high dimension data. There are lots of intercept and slope satisfied multiple linear regression Y=b0+b1x1+b2x2+b3x3
But only one of them is nearest center(has the smallest NORM), and SAS use OLS or ML to get that one .