Timeline for What is the difference between linear regression on y with x and x with y?
Current License: CC BY-SA 4.0
8 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 4 at 8:41 | comment | added | Linus Fernandes | Wouldn't that be minimize $ \frac{-1} {b^{2}}$? | |
| Feb 10, 2019 at 13:01 | comment | added | whuber♦ | +1: you have clearly made your point now! | |
| Feb 10, 2019 at 12:43 | history | edited | SiXUlm | CC BY-SA 4.0 |
added 92 characters in body
|
| Feb 10, 2019 at 12:42 | comment | added | SiXUlm | @AustinShin actually I cheated a bit here. In the middle line, I take out $b$, then change variable: $b: = 1/b$, which then gives me the last line. | |
| Feb 7, 2019 at 2:58 | comment | added | Austin Shin | how is the last line equivalent to the middle line? If you multiply out 1/b^2 you get E(X - Y/b)^2 not E(X - Yb)^2 | |
| Dec 15, 2018 at 11:35 | comment | added | SiXUlm | You are correct. Actually I thought about it but could not find a simple (and less mathematical) way to explain why two solutions are necessarily different, that's why I tried to make to these two problems $\textit{look}$ as similar as possible. Here, I'm just trying to provide a different viewpoint. | |
| Dec 14, 2018 at 23:25 | comment | added | whuber♦ | Although this is correct--and thank you for these observations--you leave your readers hanging: could you explain why the solutions to these two different looking problems are necessarily different? | |
| Dec 14, 2018 at 21:06 | history | answered | SiXUlm | CC BY-SA 4.0 |