hex.schemas.RuleFitV3 Maven / Gradle / Ivy
package hex.schemas;
import hex.rulefit.RuleFit;
import hex.rulefit.RuleFitModel;
import water.api.API;
import water.api.schemas3.ModelParametersSchemaV3;
public class RuleFitV3 extends ModelBuilderSchema {
public static final class RuleFitParametersV3 extends ModelParametersSchemaV3 {
public static final String[] fields = new String[] {
"model_id",
"training_frame",
"validation_frame",
"seed",
"response_column",
"ignored_columns",
"algorithm",
"min_rule_length",
"max_rule_length",
"max_num_rules",
"model_type",
"weights_column",
"distribution",
"rule_generation_ntrees",
"auc_type",
"remove_duplicates",
"lambda",
"max_categorical_levels",
};
@API(help = "Seed for pseudo random number generator (if applicable).", gridable = true)
public long seed;
// Input fields
@API(help = "The algorithm to use to generate rules.",
values = {"AUTO", "DRF", "GBM"})
public RuleFitModel.Algorithm algorithm;
@API(help = "Minimum length of rules. Defaults to 3.")
public int min_rule_length;
@API(help = "Maximum length of rules. Defaults to 3.")
public int max_rule_length;
@API(help = "The maximum number of rules to return. defaults to -1 which means the number of rules is selected \n" +
"by diminishing returns in model deviance.")
public int max_num_rules;
@API(help = "Specifies type of base learners in the ensemble.", values = {"RULES_AND_LINEAR", "RULES", "LINEAR"})
public RuleFitModel.ModelType model_type;
@API(help = "Specifies the number of trees to build in the tree model. Defaults to 50.")
public int rule_generation_ntrees;
@API(help = "Whether to remove rules which are identical to an earlier rule. Defaults to true." )
public boolean remove_duplicates;
@API(help = "Lambda for LASSO regressor.")
public double[] lambda;
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy