Density Mixed Data Kernel Learner

Density estimator for discrete and continuous variables. Calls np::npudens() from np.

Dictionary

This Learner can be instantiated via lrn():

lrn("dens.mixed")

Meta Information

• Task type: “dens”

• Predict Types: “pdf”

• Feature Types: “integer”, “numeric”

• Required Packages: mlr3, mlr3proba, mlr3extralearners, np

Parameters

Id	Type	Default	Levels	Range
bws	untyped	-		-
ckertype	character	gaussian	gaussian, epanechnikov, uniform	-
bwscaling	logical	FALSE	TRUE, FALSE	-
bwmethod	character	cv.ml	cv.ml, cv.ls, normal-reference	-
bwtype	character	fixed	fixed, generalized_nn, adaptive_nn	-
bandwidth.compute	logical	FALSE	TRUE, FALSE	-
ckerorder	integer	2		\([2, 8]\)
remin	logical	TRUE	TRUE, FALSE	-
itmax	integer	10000		\([1, \infty)\)
nmulti	integer	-		\([1, \infty)\)
ftol	numeric	1.490116e-07		\((-\infty, \infty)\)
tol	numeric	0.0001490116		\((-\infty, \infty)\)
small	numeric	1.490116e-05		\((-\infty, \infty)\)
lbc.dir	numeric	0.5		\((-\infty, \infty)\)
dfc.dir	numeric	0.5		\((-\infty, \infty)\)
cfac.dir	untyped	2.5 * (3 - sqrt(5))		-
initc.dir	numeric	1		\((-\infty, \infty)\)
lbd.dir	numeric	0.1		\((-\infty, \infty)\)
hbd.dir	numeric	1		\((-\infty, \infty)\)
dfac.dir	untyped	0.25 * (3 - sqrt(5))		-
initd.dir	numeric	1		\((-\infty, \infty)\)
lbc.init	numeric	0.1		\((-\infty, \infty)\)
hbc.init	numeric	2		\((-\infty, \infty)\)
cfac.init	numeric	0.5		\((-\infty, \infty)\)
lbd.init	numeric	0.1		\((-\infty, \infty)\)
hbd.init	numeric	0.9		\((-\infty, \infty)\)
dfac.init	numeric	0.37		\((-\infty, \infty)\)
ukertype	character	-	aitchisonaitken, liracine	-
okertype	character	-	wangvanryzin, liracine	-

References

Li, Qi, Racine, Jeff (2003). “Nonparametric estimation of distributions with categorical and continuous data.” journal of multivariate analysis, 86(2), 266–292.

See also

• Dictionary of Learners: mlr3::mlr_learners.

• as.data.table(mlr_learners) for a table of available Learners in the running session (depending on the loaded packages).

• Chapter in the mlr3book: https://mlr3book.mlr-org.com/basics.html#learners

• mlr3learners for a selection of recommended learners.

• mlr3cluster for unsupervised clustering learners.

• mlr3pipelines to combine learners with pre- and postprocessing steps.

• mlr3tuning for tuning of hyperparameters, mlr3tuningspaces for established default tuning spaces.

Author

RaphaelS1

Super classes

mlr3::Learner -> mlr3proba::LearnerDens -> LearnerDensMixed

Methods

Public methods

• LearnerDensMixed$new()

• LearnerDensMixed$clone()

Inherited methods

• mlr3::Learner$base_learner()
• mlr3::Learner$configure()
• mlr3::Learner$encapsulate()
• mlr3::Learner$format()
• mlr3::Learner$help()
• mlr3::Learner$predict()
• mlr3::Learner$predict_newdata()
• mlr3::Learner$print()
• mlr3::Learner$reset()
• mlr3::Learner$selected_features()
• mlr3::Learner$train()

---

Method new()

Creates a new instance of this R6 class.

Usage

LearnerDensMixed$new()

---

Method clone()

The objects of this class are cloneable with this method.

Usage

LearnerDensMixed$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

# Define the Learner
learner = mlr3::lrn("dens.mixed")
print(learner)
#> 
#> ── <LearnerDensMixed> (dens.mixed): Kernel Density Estimator ───────────────────
#> • Model: -
#> • Parameters: list()
#> • Packages: mlr3, mlr3proba, mlr3extralearners, and np
#> • Predict Types: [pdf]
#> • Feature Types: integer and numeric
#> • Encapsulation: none (fallback: -)
#> • Properties:
#> • Other settings: use_weights = 'error'

# Define a Task
task = mlr3::tsk("faithful")

# Create train and test set
ids = mlr3::partition(task)

# Train the learner on the training ids
learner$train(task, row_ids = ids$train)

print(learner$model)
#> MixedKDE_gaussian() 


# Make predictions for the test rows
predictions = learner$predict(task, row_ids = ids$test)
#> 
Multistart 1 of 1 |
Multistart 1 of 1 |
Multistart 1 of 1 |
Multistart 1 of 1 /
Multistart 1 of 1 |
Multistart 1 of 1 |
                   


# Score the predictions
predictions$score()
#> dens.logloss 
#>    0.9930027