0
$\begingroup$

I have two models.

Model 1: I have a dataset of American high school students and their test scores and other characteristics. I built an ARDRegression model that predicts how well a student will perform in an American based college based on their high school data.

Model 2: I have a dataset of high school students around the entire world (including the American students used in the first model). I translated the test scores for the international students to the American high school level courses using a Network. I built an ARDRegression model that predicts how well all students will perform in an American based college.

Model 2 is not as good at projecting American high school student success in college as the first model. In general, both models value the same input variables, but the first model has specific categorical inputs strictly for American high school students which makes the first model better at projecting American high school students.

I want to combine the results of both models, so the American high school student projection is a combination of the output from Model 1 and Model 2.

I was thinking of doing a weighted average of the American high school student results, but I didn't know if this would be correct.

Any ideas would be appreciated. Thanks.

$\endgroup$

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.