-
`sklearn` provides a [`BaseEnsemble`](https://github.com/scikit-learn/scikit-learn/blob/70fdc843a4b8182d97a3508c1a426acc5e87e980/sklearn/ensemble/_base.py#L82) class which can be used to ensemble vari…
-
Traceback (most recent call last):
File "main.py", line 174, in
main()
File "main.py", line 170, in main
learner.train()
File "/home/research/lanier.m/Desktop/SDVT-master/metalearn…
-
As of now, covariates `X`, treatment assignments `w` and observey outcomes `y` can be provided as `numpy` datastructures (`np.ndarray`) or as `pandas` datastructures (`pd.Series` and `pd.DataFrame`) r…
-
Several MetaLearners, such as the R-Learner or DR-Learner, have propensity base models.
As of now, they are trained -- just as all other base models -- based on the data passed through the `MetaLea…
-
Algo-param M:
- [ ] [max_abs_leafnode_pred](https://docs.h2o.ai/h2o/latest-stable/h2o-docs/data-science/algo-params/max_abs_leafnode_pred.html)
- [ ] [max_active_predictors](https://docs.h2o.ai/h2…
-
#791 ought to allow missing data when inferencing if allow_missing = True, but currently an error is thrown when the inference data includes missing data. RE: below for a working example based on the …
-
# Problem
When we use the ordinary classification models provided by sklearn, we get poor results because metalearners call the `predict` methods that return 0 or 1.
It would be because metalear…
-
Currently, we just run a single ensembler pipeline during AutoMLSearch after fitting all available model families. In this pipeline , we are using the default parameters for the metalearner.
When …
-
Currently the metalearners (`SLearner`, `TLearner` and `XLearner`) only accept Y, T and X variables. So one cannot differentiate between X and W. All these methods can be easily adapted to handle X an…
-
Here is the initial design idea for the Stacked Ensemble grid search (which is mostly a search of the metalearner hyperparameters, but could also include other params like metalearner_nfolds).
Py…