cv_fit_cate: Cross-Validated Fitting of the Conditional Average Treatment...
In Netflix/sherlock: Causal Machine Learning for Segment Discovery and Analysis

Estimation of the conditional average treatment effect (CATE), either within particular folds induced by cross-validation sample-splitting or upon the full data. In the latter case, a copy of the full data is made (which is NOT recommended). Whether the CATE is estimated within folds or not is specified by the split_type argument of est_cate. Estimation of the CATE is performed by computing the conditional mean of the doubly robust transformed pseudo-outcome on the specified set of segmenation covariates.

1	cv_fit_cate(fold = NULL, data_for_cate, segment_by, cate_learner)

`fold`	An object specifying the cross-validation folds into which the observations fall, as generated by `make_folds`.
`data_for_cate`	A `data.table` with additional columns containing nuisance parameter estimates (propensity score, outcome regression, doubly robust transformed pseudo-outcome) appended. A dataset of this form can be used directly for estimation of the CATE.
`segment_by`	A `character` vector specifying the column names in `data_obs` that correspond to the covariates over which segmentation should be performed. This should be a strict subset of `baseline`.
`cate_learner`	An instantiated learner object, with class inheriting from `Lrnr_base`, from sl3, to be used to estimate the CATE, based on a regression of a doubly robust pseudo-outcome on the specified segmentation covariates. Note that the outcome of this estimation task is derived from the other nuisance parameter estimates and should be expected to always be continuous-valued, so algorithms or ensembles should be set up accordingly.

A list (as required by cross_validate) containing a trained sl3 learner object in its first slot and the predicted values of the CATE in its second slot.

Netflix/sherlock documentation built on Dec. 17, 2021, 5:22 a.m.