Least-Squares Fit: Find Value of B for (-1,2), (0,1), (3,-4)

[JT73]

Given the data points (-1,2) (0,1) and (3,-4), you want to fit the best straight line, y=mx + b, through these points. given that m = -20/13, find the value of b that minimizes the error E = Ʃ ("N" on top, "i=1" on bottom) (MXi + B - Yi)^2.

This was on a test I just took for calc 1. I did the work there, I am not going to put it all on here via computer. Basically I took the derivative of the function with respect to B. Then did three iterations of it using the data points. Then solved for B. I got b = .691. Can anyone confirm/refute this?


Thanks

 

[Mark44]

You're asking a lot for helpers to work the whole problem for you and compare what they get to your answer. Forum requirements are that you show what you've done.

 

[Dick]

Given the data points (-1,2) (0,1) and (3,-4), you want to fit the best straight line, y=mx + b, through these points. given that m = -20/13, find the value of b that minimizes the error E = Ʃ ("N" on top, "i=1" on bottom) (MXi + B - Yi)^2.

This was on a test I just took for calc 1. I did the work there, I am not going to put it all on here via computer. Basically I took the derivative of the function with respect to B. Then did three iterations of it using the data points. Then solved for B. I got b = .691. Can anyone confirm/refute this?Thanks

Yes, it's pretty close. If you work a little harder you can get an exact fractional answer for B like the -20/13 for the slope.