fastCV: The Fast Cross-Validation via Sequential Testing (CVST)...

Description Usage Arguments Value Author(s) References See Also Examples

View source: R/CV.R

Description

CVST is an improved cross-validation procedure which uses non-parametric testing coupled with sequential analysis to determine the best parameter set on linearly increasing subsets of the data. By eliminating underperforming candidates quickly and keeping promising candidates as long as possible, the method speeds up the computation while preserving the capability of a full cross-validation.

Usage

1
fastCV(train, learner, params, setup, test = NULL, verbose = TRUE)

Arguments

train

The data set as CVST.data object.

learner

The learner as CVST.learner object.

params

the parameter grid as CVST.params object.

setup

A CVST.setup object containing the necessary parameter for the CVST procedure.

test

An independent test set that should be used at each step. If NULL then the remaining data after learning a model at each step is used instead.

verbose

Should the procedure report the performance after each step?

Value

Returns the optimal parameter settings as determined by fast cross-validation via sequential testing.

Author(s)

Tammo Krueger <tammokrueger@googlemail.com>

References

Tammo Krueger, Danny Panknin, and Mikio Braun. Fast cross-validation via sequential testing. Journal of Machine Learning Research 16 (2015) 1103-1155. URL http://jmlr.org/papers/volume16/krueger15a/krueger15a.pdf.

See Also

CV constructCVSTModel constructData constructLearner constructParams

Examples

1
2
3
4
ns = noisySine(100)
svm = constructSVMLearner()
params = constructParams(kernel="rbfdot", sigma=10^(-3:3), nu=c(0.05, 0.1, 0.2, 0.3))
opt = fastCV(ns, svm, params, constructCVSTModel())

Example output

Loading required package: kernlab
Loading required package: Matrix
Total number of params: 28 
(sim: 26 alpha: 0.05 left: 28 )(sim: 14 alpha: 0.05 left: 28 )Skipped configurations: 0  (sim: 14 alpha: 0.05 left: 28 )Skipped configurations: 2  (sim: 12 alpha: 0.05 left: 26 )Skipped configurations: 2  (sim: 9 alpha: 0.05 left: 26 )Skipped configurations: 14  EARLY STOPPING!

CVST documentation built on May 2, 2019, 3:43 p.m.