Could You Have Stopped This? Vegas Shooting Analysis with Attorneys on Retainer

[ARCHIVED THREAD] - Quadratic Regression

Posted: 12/4/2010 3:47:27 PM EDT

I am given a table

523-0
563-1
643-2
703-3
698-4
648-5
618-6

I listed the y first and then the x. I have tried for about an hour, I got it graphed via a quadratic regression model and it gave me this which is wrong.

13.45x^298.93x+506.11

Whatever answer I get I am supposed to round to the nearest hundreth which I have done.

Posted: 12/4/2010 3:55:44 PM EDT

[#1]

How do you know the answer is wrong? Curve fitting is about minimizing error, but doesn't necessarily give a perfect fit to the data.

Posted: 12/4/2010 3:56:29 PM EDT

[#2]

Posted: 12/4/2010 4:22:26 PM EDT

[#3]

Quoted:
http://media.ar15.com/media/viewFile.html?i=21717

Aimless poop thread?

Posted: 12/4/2010 4:27:55 PM EDT

[#4]

x^x is not quadratic. You seem to be running the wrong regression. Are you doing this by hand or using a computer? I got y = 573.36+18.21x by computer, I don't have time to do it by hand.

Posted: 12/4/2010 4:56:32 PM EDT

[#5]

Regression I get is:

-13.452x^2 + 98.929x +506.1 and this is with a R-square value of .9262

Posted: 12/4/2010 5:12:52 PM EDT

[#6]

More likely you're supposed to use least squares regression. That will give you a function that describes a straight line, which sorta-kinda describes the plot of the points you were given. Can you give us the exact wording of the problem?

A quadratic equation will give you a line that is parabola-shaped.

Posted: 12/4/2010 5:19:31 PM EDT

[#7]

A plot shows that the data will not fit a quadratic equation precisely; there are three inflections.

Your fit is as good as it gets if the rule is a quadratic fit. A fourth order equation almost fits.

[ARCHIVED THREAD] - Quadratic Regression

Warning

Confirm Action

[ARCHIVED THREAD] - Quadratic Regression

[ARCHIVED THREAD] - Quadratic Regression