^fitting I need to stop tweeting from my phone
Show this thread
I just realized (with help) that when the White House talks about a cubic fit they mean cubic splines, not feeling a third-degree polynomial. But holy smokes are they really using that for prediction?https://twitter.com/WhiteHouseCEA45/status/1257680258364555264 …
-
-
Thanks. Twitter will use this info to make your timeline better. UndoUndo
-
-
-
not a big fan of economists playing epidemiologist?
Thanks. Twitter will use this info to make your timeline better. UndoUndo
-
-
-
Dangers of extrapolation and the blight of Dunning-Kruger....
-
This Tweet is unavailable.
- Show replies
New conversation -
-
-
Can you explain a cubic spline & issues with it?
-
A cubic spline comprises cubic polynomials smoothly joined together to fit data. They've a smoothing parameter that can be adjusted to fit/overfit the data more closely. I've never used them to extrapolate. Like cubic polys, I imagine that they leap to +/- ∞ when extrapolating.
- Show replies
New conversation -
-
-
You sure that it's a cubic spline and not, say, a 3rd-degree fit to log(daily deaths)? Because I don't see how do you extrapolate a spline.
Thanks. Twitter will use this info to make your timeline better. UndoUndo
-
-
-
So, is this basically a Generalized Additive Model (GAM)? family = poisson?
-
I don't think so. Think
@CT_Bergstrom isn't correct here - doesn't look like a cubic spline to me. Looks like a plain old 3d degree polynomial fit...
End of conversation
New conversation -
-
-
Wait does this mean they're using splines to produce a series of local regressions or are they trying to fit a spline to the entire outbreak?
-
Come now. They are obviously doing the stupider... i mean latter
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.