Function for Tuning a LearnerSource:
Function to tune a mlr3::Learner.
The function internally creates a TuningInstanceSingleCrit or TuningInstanceMultiCrit which describe the tuning problem.
It executes the tuning with the Tuner (
tuner) and returns the result with the tuning instance (
The ArchiveTuning (
$archive) stores all evaluated hyperparameter configurations and performance scores.
tune( tuner, task, learner, resampling, measures = NULL, term_evals = NULL, term_time = NULL, terminator = NULL, search_space = NULL, store_benchmark_result = TRUE, store_models = FALSE, check_values = FALSE, allow_hotstart = FALSE, keep_hotstart_stack = FALSE, evaluate_default = FALSE, callbacks = list(), method )
Task to operate on.
Learner to tune.
Resampling that is used to evaluate the performance of the hyperparameter configurations. Uninstantiated resamplings are instantiated during construction so that all configurations are evaluated on the same data splits. Already instantiated resamplings are kept unchanged. Specialized Tuner change the resampling e.g. to evaluate a hyperparameter configuration on different data splits. This field, however, always returns the resampling passed in construction.
Number of allowed evaluations. Ignored if
Maximum allowed time in seconds. Ignored if
Stop criterion of the tuning process.
Hyperparameter search space. If
NULL(default), the search space is constructed from the TuneToken of the learner's parameter set (learner$param_set).
TRUE(default), store resample result of evaluated hyperparameter configurations in archive as mlr3::BenchmarkResult.
TRUE, fitted models are stored in the benchmark result (
store_benchmark_result = FALSE, models are only stored temporarily and not accessible after the tuning. This combination is needed for measures that require a model.
TRUE, hyperparameter values are checked before evaluation and performance scores after. If
FALSE(default), values are unchecked but computational overhead is reduced.
Allow to hotstart learners with previously fitted models. See also mlr3::HotstartStack. The learner must support hotstarting. Sets
store_models = TRUE.
TRUE, mlr3::HotstartStack is kept in
TRUE, learner is evaluated with hyperparameters set to their default values at the start of the optimization.
(list of CallbackTuning)
List of callbacks.
The mlr3::Task, mlr3::Learner, mlr3::Resampling, mlr3::Measure and Terminator are used to construct a TuningInstanceSingleCrit.
If multiple performance Measures are supplied, a TuningInstanceMultiCrit is created.
term_time are shortcuts to create a Terminator.
If both parameters are passed, a TerminatorCombo is constructed.
For other Terminators, pass one with
If no termination criterion is needed, set
The search space is created from paradox::TuneToken or is supplied by
There are several sections about hyperparameter optimization in the mlr3book.
The gallery features a collection of case studies and demos about optimization.
If no measure is passed, the default measure is used. The default measure depends on the task type.
For analyzing the tuning results, it is recommended to pass the ArchiveTuning to
The returned data table is joined with the benchmark result which adds the mlr3::ResampleResult for each hyperparameter evaluation.
The archive provides various getters (e.g.
$learners()) to ease the access.
All getters extract by position (
i) or unique hash (
For a complete list of all getters see the methods section.
The benchmark result (
$benchmark_result) allows to score the hyperparameter configurations again on a different measure.
Alternatively, measures can be supplied to
The mlr3viz package provides visualizations for tuning results.
# Hyperparameter optimization on the Palmer Penguins data set task = tsk("pima") # Load learner and set search space learner = lrn("classif.rpart", cp = to_tune(1e-04, 1e-1, logscale = TRUE) ) # Run tuning instance = tune( tuner = tnr("random_search", batch_size = 2), task = tsk("pima"), learner = learner, resampling = rsmp ("holdout"), measures = msr("classif.ce"), terminator = trm("evals", n_evals = 4) ) # Set optimal hyperparameter configuration to learner learner$param_set$values = instance$result_learner_param_vals # Train the learner on the full data set learner$train(task) # Inspect all evaluated configurations as.data.table(instance$archive) #> cp classif.ce x_domain_cp runtime_learners timestamp #> 1: -6.359271 0.2500000 0.001730628 0.011 2023-06-27 06:09:48 #> 2: -3.339324 0.2929688 0.035460906 0.011 2023-06-27 06:09:48 #> 3: -4.842755 0.2460938 0.007885300 0.011 2023-06-27 06:09:48 #> 4: -6.128559 0.2500000 0.002179720 0.012 2023-06-27 06:09:48 #> batch_nr warnings errors resample_result #> 1: 1 0 0 <ResampleResult> #> 2: 1 0 0 <ResampleResult> #> 3: 2 0 0 <ResampleResult> #> 4: 2 0 0 <ResampleResult>