resume

Resume training of cross-validated regression ensemble model

Syntax

ens1 = resume(ens,nlearn)

ens1 = resume(ens,nlearn,Name=Value)

Description

ens1 = resume(ens,nlearn) continues training in every fold with the same options used to train ens, except parallel training options and printout frequency, for nlearn more training cycles. The function returns a new cross-validated regression ensemble model ens1.

ens1 = resume(ens,nlearn,Name=Value) specifies additional options using one or more name-value arguments. For example, you can specify the printout frequency, and set options for computing in parallel.

Examples

Cross-Validate Regression Ensemble Augmented with Additional Training

Examine the cross-validation error after training a regression ensemble for more cycles.

Load the carsmall data set and select displacement, horsepower, and vehicle weight as predictors.

load carsmall
X = [Displacement Horsepower Weight];

Train a regression ensemble for 50 cycles.

ens = fitrensemble(X,MPG,'NumLearningCycles',50); 

Cross-validate the ensemble and examine the cross-validation error.

rng(10,'twister') % For reproducibility
cvens = crossval(ens);
L = kfoldLoss(cvens)

L = 
27.9435

Train for 50 more cycles and examine the new cross-validation error.

cvens = resume(cvens,50);
L = kfoldLoss(cvens)

L = 
28.7114

The additional training did not improve the cross-validation error.

Input Arguments

ens — Cross-validated regression ensemble model

RegressionPartitionedEnsemble model object

Cross-validated regression ensemble model, specified as a RegressionPartitionedEnsemble model object created with one of these functions:

• fitrensemble with one of these five cross-validation name-value argument CrossVal, KFold, Holdout, Leaveout, or CVPartition

• crossval applied to a RegressionEnsemble model object

nlearn — Number of additional training cycles

positive integer

Number of additional training cycles for ens, specified as a positive integer.

Data Types: double | single

Name-Value Arguments

Specify optional pairs of arguments as Name1=Value1,...,NameN=ValueN, where Name is the argument name and Value is the corresponding value. Name-value arguments must appear after other arguments, but the order of the pairs does not matter.

Before R2021a, use commas to separate each name and value, and enclose Name in quotes.

Example: resume(ens,10,NPrint=5,Options=statset(UseParallel=true)) specifies to train ens for an additional 10 cycles, display a message to the command line every time resume finishes training 5 folds, and perform computations in parallel.

NPrint — Printout frequency

"off" (default) | positive integer

Printout frequency, specified as a positive integer m or "off". resume displays a message to the command line every time it finishes training m folds. If you specify "off", resume does not display a message when it completes training folds.

Tip

For the fastest training of some boosted decision trees, when the regression method is "LSBoost", set NPrint to "off" (the default value).

Example: NPrint=5

Data Types: single | double | char | string

Options — Options for computing in parallel and setting random number streams

structure

Options for computing in parallel and setting random number streams, specified as a structure. Create the Options structure using statset.

Note

You need Parallel Computing Toolbox™ to run computations in parallel.

You can use the same parallel options for resume as you used for the original training. Use the Options argument to change the parallel options, as needed. This table describes the option fields and their values.

Field Name	Value	Default
UseParallel	Set this value to true to compute in parallel. Parallel ensemble training requires you to set the Method name-value argument to "Bag". Parallel training is available only for tree learners, the default type for Method="Bag".	false
UseSubstreams	Set this value to true to perform computations in a reproducible manner. To compute reproducibly, set Streams to a type that allows substreams: "mlfg6331_64" or "mrg32k3a".	false
Streams	Specify this value as a RandStream object or cell array of such objects. Use a single object except when the UseParallel value is true and the UseSubstreams value is false. In that case, use a cell array that has the same size as the parallel pool.	If you do not specify Streams, resume uses the default stream or streams.

For dual-core systems and above, resume parallelizes training using Intel^® Threading Building Blocks (TBB). Therefore, setting UseParallel to true might not provide a significant increase in speed on a single computer. For details on Intel TBB, see https://www.intel.com/content/www/us/en/developer/tools/oneapi/onetbb.html.

Example: Options=statset(UseParallel=true)

Data Types: struct

Extended Capabilities

Automatic Parallel Support

Accelerate code by automatically running computation in parallel using Parallel Computing Toolbox™.

resume supports parallel training using the 'Options' name-value argument. Create options using statset, such as options = statset('UseParallel',true). Parallel ensemble training requires you to set the 'Method' name-value argument to 'Bag'. Parallel training is available only for tree learners, the default type for 'Bag'.

GPU Arrays

Accelerate code by running on a graphics processing unit (GPU) using Parallel Computing Toolbox™.

This function fully supports GPU arrays. For more information, see Run MATLAB Functions on a GPU (Parallel Computing Toolbox).

Version History

Introduced in R2012b