I am trying to use sparklyr ml_cross_validator() function to do the cross validation with machine learning models. It seems the
argument has to be specified and it is veried for different models. I tried ?ml_cross_validator, it says :
estimator_param_maps A named list of stages and hyper-parameter sets to tune. See details.
whereas the "details" says:
ml_cross_validator() performs k-fold cross validation while
ml_train_validation_split() performs tuning on one pair of train and validation datasets.
Is there any example that I can follow on how to set this argument for different models? for instance, people use
estimator_param_maps = list(logistic_regression = list(elastic_net_param = 0)) (see https://www.eddjberry.com/post/2018-12-12-sparklyr-feature-selection/) for the logistic regression model.
what is this "elastic_net_param" ? and what would be the parameter to specify here for linear regression, decision tree, random forest, and so on?
Thank you very much.