Hyperparameter Tuning with Covariance Matrix Adaptation Evolution Strategy

Subclass for Covariance Matrix Adaptation Evolution Strategy (CMA-ES). Calls adagio::pureCMAES() from package adagio.

Source

Hansen N (2016). “The CMA Evolution Strategy: A Tutorial.” 1604.00772.

Dictionary

This Tuner can be instantiated with the associated sugar function tnr():

tnr("cmaes")

Control Parameters

start_values

character(1)
Create random start values or based on center of search space? In the latter case, it is the center of the parameters before a trafo is applied.

For the meaning of the control parameters, see adagio::pureCMAES(). Note that we have removed all control parameters which refer to the termination of the algorithm and where our terminators allow to obtain the same behavior.

Progress Bars

$optimize() supports progress bars via the package progressr combined with a Terminator. Simply wrap the function in progressr::with_progress() to enable them. We recommend to use package progress as backend; enable with progressr::handlers("progress").

Logging

All Tuners use a logger (as implemented in lgr) from package bbotk. Use lgr::get_logger("bbotk") to access and control the logger.

Optimizer

This Tuner is based on bbotk::OptimizerCmaes which can be applied on any black box optimization problem. See also the documentation of bbotk.

Resources

There are several sections about hyperparameter optimization in the mlr3book.

• Learn more about tuners.

The gallery features a collection of case studies and demos about optimization.

• Use the Hyperband optimizer with different budget parameters.

See also

Other Tuner: mlr_tuners_design_points, mlr_tuners_gensa, mlr_tuners_grid_search, mlr_tuners_irace, mlr_tuners_nloptr, mlr_tuners_random_search, mlr_tuners

Super classes

mlr3tuning::Tuner -> mlr3tuning::TunerFromOptimizer -> TunerCmaes

Methods

Public methods

• TunerCmaes$new()

• TunerCmaes$clone()

Inherited methods

• mlr3tuning::Tuner$format()
• mlr3tuning::Tuner$help()
• mlr3tuning::Tuner$print()
• mlr3tuning::TunerFromOptimizer$optimize()

---

Method new()

Creates a new instance of this R6 class.

Usage

TunerCmaes$new()

---

Method clone()

The objects of this class are cloneable with this method.

Usage

TunerCmaes$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

# Hyperparameter Optimization

# load learner and set search space
learner = lrn("classif.rpart",
  cp = to_tune(1e-04, 1e-1, logscale = TRUE),
  minsplit = to_tune(p_dbl(2, 128, trafo = as.integer)),
  minbucket = to_tune(p_dbl(1, 64, trafo = as.integer))
)

# run hyperparameter tuning on the Palmer Penguins data set
instance = tune(
  tuner = tnr("cmaes"),
  task = tsk("penguins"),
  learner = learner,
  resampling = rsmp("holdout"),
  measure = msr("classif.ce"),
  term_evals = 10)

# best performing hyperparameter configuration
instance$result
#>           cp minsplit minbucket learner_param_vals  x_domain classif.ce
#> 1: -7.652763        2         1          <list[4]> <list[3]> 0.02608696

# all evaluated hyperparameter configuration
as.data.table(instance$archive)
#>            cp   minsplit minbucket classif.ce  x_domain_cp x_domain_minsplit
#>  1: -7.113395 128.000000  50.25548 0.04347826 0.0008141266               128
#>  2: -6.365598  90.335303  64.00000 0.12173913 0.0017197126                90
#>  3: -7.652763   2.000000   1.00000 0.02608696 0.0004747304                 2
#>  4: -9.210340  31.302687  43.47549 0.04347826 0.0001000000                31
#>  5: -6.097083 128.000000  64.00000 0.12173913 0.0022494200               128
#>  6: -3.833374  38.466866  64.00000 0.12173913 0.0216364922                38
#>  7: -6.841480 116.869416  64.00000 0.12173913 0.0010685208               116
#>  8: -2.302585  95.700263  24.40318 0.04347826 0.1000000000                95
#>  9: -2.302585   7.572852  36.87330 0.04347826 0.1000000000                 7
#> 10: -7.132403  21.146453   1.00000 0.02608696 0.0007987972                21
#>     x_domain_minbucket runtime_learners           timestamp batch_nr warnings
#>  1:                 50            0.007 2023-03-09 07:19:00        1        0
#>  2:                 64            0.008 2023-03-09 07:19:00        2        0
#>  3:                  1            0.009 2023-03-09 07:19:00        3        0
#>  4:                 43            0.008 2023-03-09 07:19:00        4        0
#>  5:                 64            0.008 2023-03-09 07:19:00        5        0
#>  6:                 64            0.008 2023-03-09 07:19:00        6        0
#>  7:                 64            0.011 2023-03-09 07:19:00        7        0
#>  8:                 24            0.008 2023-03-09 07:19:00        8        0
#>  9:                 36            0.008 2023-03-09 07:19:00        9        0
#> 10:                  1            0.008 2023-03-09 07:19:01       10        0
#>     errors      resample_result
#>  1:      0 <ResampleResult[21]>
#>  2:      0 <ResampleResult[21]>
#>  3:      0 <ResampleResult[21]>
#>  4:      0 <ResampleResult[21]>
#>  5:      0 <ResampleResult[21]>
#>  6:      0 <ResampleResult[21]>
#>  7:      0 <ResampleResult[21]>
#>  8:      0 <ResampleResult[21]>
#>  9:      0 <ResampleResult[21]>
#> 10:      0 <ResampleResult[21]>

# fit final model on complete data set
learner$param_set$values = instance$result_learner_param_vals
learner$train(tsk("penguins"))