Build a RuleFit Model

Builds a Distributed RuleFit model on a parsed dataset, for regression or classification.

h2o.rulefit(
  x,
  y,
  training_frame,
  model_id = NULL,
  validation_frame = NULL,
  seed = -1,
  algorithm = c("AUTO", "DRF", "GBM"),
  min_rule_length = 3,
  max_rule_length = 3,
  max_num_rules = -1,
  model_type = c("rules_and_linear", "rules", "linear"),
  weights_column = NULL,
  distribution = c("AUTO", "bernoulli", "multinomial", "gaussian", "poisson", "gamma",
    "tweedie", "laplace", "quantile", "huber"),
  rule_generation_ntrees = 50,
  auc_type = c("AUTO", "NONE", "MACRO_OVR", "WEIGHTED_OVR", "MACRO_OVO", "WEIGHTED_OVO"),
  remove_duplicates = TRUE,
  lambda = NULL,
  max_categorical_levels = 10
)

Arguments

x

(Optional) A vector containing the names or indices of the predictor variables to use in building the model. If x is missing, then all columns except y are used.

y

The name or column index of the response variable in the data. The response must be either a numeric or a categorical/factor variable. If the response is numeric, then a regression model will be trained, otherwise it will train a classification model.

training_frame

Id of the training data frame.

model_id

Destination id for this model; auto-generated if not specified.

validation_frame

Id of the validation data frame.

seed

Seed for random numbers (affects certain parts of the algo that are stochastic and those might or might not be enabled by default). Defaults to -1 (time-based random number).

algorithm

The algorithm to use to generate rules. Must be one of: "AUTO", "DRF", "GBM". Defaults to AUTO.

min_rule_length

Minimum length of rules. Defaults to 3.

max_rule_length

Maximum length of rules. Defaults to 3.

max_num_rules

The maximum number of rules to return. defaults to -1 which means the number of rules is selected by diminishing returns in model deviance. Defaults to -1.

model_type

Specifies type of base learners in the ensemble. Must be one of: "rules_and_linear", "rules", "linear". Defaults to rules_and_linear.

weights_column

Column with observation weights. Giving some observation a weight of zero is equivalent to excluding it from the dataset; giving an observation a relative weight of 2 is equivalent to repeating that row twice. Negative weights are not allowed. Note: Weights are per-row observation weights and do not increase the size of the data frame. This is typically the number of times a row is repeated, but non-integer values are supported as well. During training, rows with higher weights matter more, due to the larger loss function pre-factor. If you set weight = 0 for a row, the returned prediction frame at that row is zero and this is incorrect. To get an accurate prediction, remove all rows with weight == 0.

distribution

Distribution function Must be one of: "AUTO", "bernoulli", "multinomial", "gaussian", "poisson", "gamma", "tweedie", "laplace", "quantile", "huber". Defaults to AUTO.

rule_generation_ntrees

Specifies the number of trees to build in the tree model. Defaults to 50. Defaults to 50.

auc_type

Set default multinomial AUC type. Must be one of: "AUTO", "NONE", "MACRO_OVR", "WEIGHTED_OVR", "MACRO_OVO", "WEIGHTED_OVO". Defaults to AUTO.

remove_duplicates

Logical. Whether to remove rules which are identical to an earlier rule. Defaults to true. Defaults to TRUE.

lambda

Lambda for LASSO regressor.

max_categorical_levels

For every categorical feature, only use this many most frequent categorical levels for model training. Only used for categorical_encoding == EnumLimited. Defaults to 10.

Examples

if (FALSE) { # \dontrun{
library(h2o)
h2o.init()

# Import the titanic dataset:
f <- "https://s3.amazonaws.com/h2o-public-test-data/smalldata/gbm_test/titanic.csv"
coltypes <- list(by.col.name = c("pclass", "survived"), types=c("Enum", "Enum"))
df <- h2o.importFile(f, col.types = coltypes)

# Split the dataset into train and test
splits <- h2o.splitFrame(data = df, ratios = 0.8, seed = 1)
train <- splits[[1]]
test <- splits[[2]]

# Set the predictors and response; set the factors:
response <- "survived"
predictors <- c("age", "sibsp", "parch", "fare", "sex", "pclass")

# Build and train the model:
rfit <- h2o.rulefit(y = response,
                    x = predictors,
                    training_frame = train,
                    max_rule_length = 10,
                    max_num_rules = 100,
                    seed = 1)

# Retrieve the rule importance:
print(rfit@model$rule_importance)

# Predict on the test data:
h2o.predict(rfit, newdata = test)
} # }