Classification Neural Network Learner

Simple and fast neural nets for tabular data classification. Calls fastai::tabular_learner() from package fastai.

Installation

The Python dependencies are automatically installed via reticulate::py_require(). See Installing Python Dependencies for more details. You can manually specify a virtual environment by calling reticulate::use_virtualenv() prior to calling the $train() function. In this virtual environment, the fastai package and its dependencies must be installed.

Saving a Learner

In order to save a lrn("classif.fastai") for later usage, it is necessary to call the $marshal() method on the Learner before writing it to disk, as the object will otherwise not be saved correctly. After loading a marshaled lrn("classif.fastai") into R again, you then need to call $unmarshal() to transform it into a useable state.

Initial parameter values

• n_epoch: Needs to be set for fastai::fit() to work. If no value is given, it is set to 5.

• eval_metric: Needs to be set for fastai::predict() to work. If no value is given, it is set to fastai::accuracy().

Dictionary

This Learner can be instantiated via lrn():

lrn("classif.fastai")

Meta Information

• Task type: “classif”

• Predict Types: “response”, “prob”

• Feature Types: “logical”, “integer”, “numeric”, “factor”, “ordered”

• Required Packages: mlr3, mlr3extralearners, fastai, reticulate

Parameters

Id	Type	Default	Levels	Range
act_cls	untyped	-		-
bn_cont	logical	TRUE	TRUE, FALSE	-
bn_final	logical	FALSE	TRUE, FALSE	-
drop_last	logical	FALSE	TRUE, FALSE	-
embed_p	numeric	0		\([0, 1]\)
emb_szs	untyped	NULL		-
n_epoch	integer	5		\([1, \infty)\)
eval_metric	untyped	-		-
layers	untyped	-		-
loss_func	untyped	-		-
lr	numeric	0.001		\([0, \infty)\)
metrics	untyped	-		-
n_out	integer	-		\((-\infty, \infty)\)
num_workers	integer	-		\((-\infty, \infty)\)
opt_func	untyped	-		-
patience	integer	1		\([1, \infty)\)
pin_memory	logical	TRUE	TRUE, FALSE	-
procs	untyped	NULL		-
ps	untyped	NULL		-
shuffle	logical	FALSE	TRUE, FALSE	-
train_bn	logical	TRUE	TRUE, FALSE	-
wd	integer	-		\([0, \infty)\)
wd_bn_bias	logical	FALSE	TRUE, FALSE	-
use_bn	logical	TRUE	TRUE, FALSE	-
y_range	untyped	NULL		-
bs	integer	50		\((-\infty, \infty)\)

References

Howard, Jeremy, Gugger, Sylvain (2020). “Fastai: A Layered API for Deep Learning.” Information, 11(2), 108. ISSN 2078-2489, doi:10.3390/info11020108 .

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

annanzrv

Super classes

mlr3::Learner -> mlr3::LearnerClassif -> LearnerClassifFastai

Active bindings

internal_valid_scores

(named list() or NULL) The validation scores extracted from eval_protocol which itself is set by fitting the fastai::tab_learner. If early stopping is activated, this contains the validation scores of the model for the optimal n_epoch, otherwise the n_epoch for the final model.

internal_tuned_values

(named list() or NULL) If early stopping is activated, this returns a list with n_epoch, which is the last epoch that yielded improvement w.r.t. the patience, extracted by max(eval_protocol$epoch)+1

validate

(numeric(1) or character(1) or NULL) How to construct the internal validation data. This parameter can be either NULL, a ratio, "test", or "predefined".

marshaled

(logical(1)) Whether the learner has been marshaled.

Methods

Public methods

• LearnerClassifFastai$new()

• LearnerClassifFastai$marshal()

• LearnerClassifFastai$unmarshal()

• LearnerClassifFastai$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$train()
• mlr3::LearnerClassif$predict_newdata_fast()

---

Method new()

Creates a new instance of this R6 class.

Usage

LearnerClassifFastai$new()

---

Method marshal()

Marshal the learner's model.

Usage

LearnerClassifFastai$marshal(...)

Arguments

...

(any)
Additional arguments passed to mlr3::marshal_model().

---

Method unmarshal()

Unmarshal the learner's model.

Usage

LearnerClassifFastai$unmarshal(...)

Arguments

...

(any)
Additional arguments passed to mlr3::marshal_model().

---

Method clone()

The objects of this class are cloneable with this method.

Usage

LearnerClassifFastai$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.