Compute Mean Square Error on test data set

John4 — Tue, 18 Apr 2017 14:38:48 GMT

I split my data into training/validation/test. I use PROC ADAPTIVEREG and i want to compute Mean Square Error on test data

This is my code :

proc adaptivereg data=test_data;
class job;
model money = age  job / dist=gamma link=log ;
run;

how can I compute Mean Square Error on this data?

Re: Compute Mean Square Error on test data set

Reeza — Tue, 18 Apr 2017 14:49:15 GMT

Use the SCORE statement to score the test data with the model and then use the output provided to calculate the Mean Square Error.