enpls.ad: Ensemble Partial Least Squares for Applicability Domain

In wind22zhu/enpls1.2: Ensemble Partial Least Squares Regression

Description

This function performs applicability domain with ensemble partial least squares.

Usage

enpls.ad(x, y, x.test = NULL, y.test = NULL, maxcomp = NULL,
  MCtimes = 500L, verbose = FALSE, method = c("mc", "bootstrap",
  "jagging"), ratio = 0.8, parallel = 1L)

Arguments

x	predictor matrix.
y	response vector.
x.test	predictor matrix for test.
y.test	response vector for test.
maxcomp	Maximum number of components included within the models, if not specified, default is the variable (column) numbers in x.
MCtimes	times of Monte-Carlo.
verbose	shall we print the MCtimes process.
method	"mc" or "bootstrap" or "jagging". Default is "mc".
ratio	sample ratio used when method = "mc" and method = "jagging".
parallel	Integer. Number of parallel processes to use. Default is 1, which means run serially.

Details

This function performs applicability domain with ensemble partial least squares.

Value

A list containing four components:

• STD.cv - STD value for training set

• STD.te - STD value for test set

• error.cv - absolute prediction error of training set

• error.te - absolute prediction error of test set

Author(s)

Min-feng Zhu <[e-mail]>, Nan Xiao <[e-mail]>

References

Kaneko H, Funatsu K. "Applicability Domain Based on Ensemble Learning in Classification and Regression Analyses." Journal of chemical information and modeling 54, no. 9 (2014): 2469-2482.

See Also

See enpls.fs for feature selection with ensemble PLS. See enpls.en for ensemble PLS regression. See enpls.od for Outlier Detection with ensemble PLS

Examples

data(logS)
x = logS$x
y = logS$y
x.test1 = logS$x.test1
y.test1 = logS$y.test1

set.seed(42)
ad_test1 = enpls.ad(x, y, x.test1, y.test1, MCtimes = 10)
print(ad_test1)
plot(ad_test1)

x.test2 = logS$x.test2
y.test2 = logS$y.test2

set.seed(42)
ad_test2 = enpls.ad(x, y, x.test2, y.test2, MCtimes = 10)
print(ad_test2)
plot(ad_test2)

wind22zhu/enpls1.2 documentation built on May 4, 2019, 6:31 a.m.