SEE vs. RMSE

CFA Level II
#1

is SEE = RMSE? MSE = SSE / n-k-1 RMSE = sqrt (MSE) SEE = sqrt (SSE / n-k-1) what say?

#2

agreed. RSS/TSS = R or R^2

#3

Yes.

#4

So SEE = RMSE. Are we saying this?

#5

Ok several websites I have found say, RMSE = sqrt ( SSE / n) whereas SEE is sqrt ( SSE / (n-k-1)) I forgot the equation but there is an example in CFAI book

#6

I think SEE is the variability of the residuals over the fitted line - i.e. it’s a standard deviation measure, but measures in-sample variability. Lower the better. RMSE is mostly used for out-of-sample predictions to see if the model is good for predicting beyond the sample space considered. Lower the better. So - they are the same mathematically. (I think so…) Missing JDV.

#7

Me too :*-(