constructCVSTModel {CVST} | R Documentation |
This is an helper object of type CVST.setup
conatining all
necessary parameters for a CVST run.
constructCVSTModel(steps = 10, beta = 0.1, alpha = 0.01, similaritySignificance = 0.05, earlyStoppingSignificance = 0.05, earlyStoppingWindow = 3, regressionSimilarityViaOutliers = FALSE)
steps |
Number of steps CVST should run |
beta |
Significance level for H0. |
alpha |
Significance level for H1. |
similaritySignificance |
Significance level of the similarity test. |
earlyStoppingSignificance |
Significance level of the early stopping test. |
earlyStoppingWindow |
Size of the early stopping window. |
regressionSimilarityViaOutliers |
Should the less strict outlier-based similarity measure for regression tasks be used. |
A CVST.setup
object suitable for fastCV
.
Tammo Krueger <tammokrueger@googlemail.com>
Tammo Krueger, Danny Panknin, and Mikio Braun. Fast cross-validation via sequential analysis. Neural Information Processing Systems (NIPS), Big Learning Workshop, 2011. URL http://biglearn.org/2011/index.php/Papers\#paper2.
Tammo Krueger, Danny Panknin, and Mikio Braun. Fast cross-validation via sequential testing. CoRR, abs/1206.2248, 2012. URL http://arxiv.org/abs/1206.2248.