H2OInfogram-class: H2OInfogram class

In h2o: R Interface for the 'H2O' Scalable Machine Learning Platform

H2OInfogram class

Description

H2OInfogram class contains a subset of what a normal H2OModel will return

Slots

model_id

string returned as part of every H2OModel

algorithm

string denoting the algorithm used to build infogram

admissible_features

string array denoting all predictor names which pass the cmi and relelvance threshold

admissible_features_valid

string array denoting all predictor names which pass the cmi and relelvance threshold from validation frame

admissible_features_xval

string array denoting all predictor names which pass the cmi and relelvance threshold from cv holdout set

net_information_threshold

numeric value denoting threshold used for predictor selection

total_information_threshold

numeric value denoting threshold used for predictor selection

safety_index_threshold

numeric value denoting threshold used for predictor selection

relevance_index_threshold

numeric value denoting threshold used for predictor selection

admissible_score

H2OFrame that contains columns, admissible, admissible_index, relevance, cmi, cmi_raw

admissible_score_valid

H2OFrame that contains columns, admissible, admissible_index, relevance, cmi, cmi_raw from validation frame

admissible_score_xval

H2OFrame that contains averages of columns, admissible, admissible_index, relevance, cmi, cmi_raw from cv hold-out

h2o documentation built on May 29, 2024, 4:26 a.m.