Stack for Hot Start Learners

This class stores learners for hot starting training, i.e. resuming or continuing from an already fitted model. We assume that hot starting is only possible if a single hyperparameter (also called the fidelity parameter, usually controlling the complexity or expensiveness) is altered and all other hyperparameters are identical.

The HotstartStack stores trained learners which can be potentially used to hot start a learner. Learner automatically hot start while training if a stack is attached to the $hotstart_stack field and the stack contains a suitable learner.

For example, if you want to train a random forest learner with 1000 trees but already have a random forest learner with 500 trees (hot start learner), you can add the hot start learner to the HotstartStack of the expensive learner with 1000 trees. If you now call the train() method (or resample() or benchmark()), a random forest with 500 trees will be fitted and combined with the 500 trees of the hotstart learner, effectively saving you to fit 500 trees.

Hot starting is only supported by learners which have the property "hotstart_forward" or "hotstart_backward". For example, an xgboost model (in mlr3learners) can hot start forward by adding more boosting iterations, and a random forest can go backwards by removing trees. The fidelity parameters are tagged with "hotstart" in learner's parameter set.

Public fields

stack

data.table::data.table()
Stores hot start learners.

hotstart_threshold

(named numeric(1))
Threshold for storing learners in the stack. If the value of the hotstart parameter is below this threshold, the learner is not added to the stack.

Methods

Public methods

• HotstartStack$new()

• HotstartStack$add()

• HotstartStack$start_cost()

• HotstartStack$format()

• HotstartStack$print()

• HotstartStack$clone()

---

Method new()

Creates a new instance of this R6 class.

Usage

HotstartStack$new(learners = NULL, hotstart_threshold = NULL)

Arguments

learners

(List of Learners)
Learners are added to the hotstart stack. If NULL (default), empty stack is created.

hotstart_threshold

(named numeric(1))
Threshold for storing learners in the stack.

---

Method add()

Add learners to hot start stack.

Usage

HotstartStack$add(learners)

Arguments

learners

(List of Learners). Learners are added to the hotstart stack.

Returns

self (invisibly).

---

Method start_cost()

Calculates the cost for each learner of the stack to hot start the target learner.

The following cost values can be returned:

• NA_real_: Learner is unsuitable to hot start target learner.

• -1: Hotstart learner in the stack and target learner are identical.

• 0 Cost for hot starting backwards is always 0.

• > 0 Cost for hot starting forward.

Usage

HotstartStack$start_cost(learner, task_hash)

Arguments

learner

Learner
Target learner.

task_hash

Task
Hash of the task on which the target learner is trained.

---

Method format()

Helper for print outputs.

Usage

HotstartStack$format(...)

Arguments

...

(ignored).

---

Method print()

Printer.

Usage

HotstartStack$print(...)

Arguments

...

(ignored).

---

Method clone()

The objects of this class are cloneable with this method.

Usage

HotstartStack$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

# train learner on pima task
task = tsk("pima")
learner = lrn("classif.debug", iter = 1)
learner$train(task)

# initialize stack with previously fitted learner
hot = HotstartStack$new(list(learner))

# retrieve learner with increased fidelity parameter
learner = lrn("classif.debug", iter = 2)

# calculate cost of hot starting
hot$start_cost(learner, task$hash)
#> [1] 1

# add stack with hot start learner
learner$hotstart_stack = hot

# train automatically uses hot start learner while fitting the model
learner$train(task)