I have learned multilevel model in Bayes methods, which I think is really cool. Different levels data could share information which could improve the total model. Is there any similar methods in non-Bayes algo, for example xgboost, which could share informations in different level data?
Asked
Active
Viewed 34 times
0
-
Something similar in the frequentist world is mixed-effects models, where a random-effects term tries to capture intra-group variation. HTH. – asifzuba Sep 03 '19 at 20:09
-
@asifzuba thanks for your key words, mixed-effects models, help me a lot! – Wendy Hu Sep 04 '19 at 06:35