This estimator combines outcome regression with inverse propensity weighting and has a second-order product bias.
Cross-fitting trains nuisance estimators outside each observation's fold and evaluates the target score on the held-out fold.
Articles by others on the same topic
There are currently no matching articles.