I am looking for pointers as to what would be the non-parametric equivalent of Nadaraya-Watson regression when modelling a binary outcome. I have been googling and ended up with Generalized Additive Models, but I think these impose functional forms?
Asked
Active
Viewed 23 times
0
-
"Nonparametric" has a pretty loose definition, but isn't Nadaraya-Watson already considered nonparametric? – Dave Apr 06 '21 at 17:18
-
@Dave, thanks for the reply. Yes, NW is non-parametric, but not suited for binary outcomes. That's why I asked what would be the binary outcome alternative. – Papayapap Apr 06 '21 at 17:22