The plsdof package provides Degrees of Freedom estimates for Partial Least Squares (PLS) Regression. Model selection for PLS is based on various information criteria (aic, bic, gmdl) or on cross-validation. Estimates for the mean and covariance of the PLS regression coefficients are available. They allow the construction of approximate confidence intervals and the application of test procedures. Further, cross-validation procedures for Ridge Regression and Principal Components Regression are available.
|Author||Nicole Kraemer, Mikio L. Braun|
|Date of publication||2014-09-04 15:41:41|
|Maintainer||Nicole Kraemer <email@example.com>|
|License||GPL (>= 2)|
benchmark.pls: Comparison of model selection criteria for Partial Least...
benchmark.regression: Comparison of Partial Least Squares Regression, Principal...
coef.plsdof: Regression coefficients
compute.lower.bound: Lower bound for the Degrees of Freedom
dA: Derivative of normalization function
dnormalize: Derivative of normalization function
dvvtz: First derivative of the projection operator
first.local.minimum: Index of the first local minimum.
information.criteria: Information criteria
kernel.pls.fit: Kernel Partial Least Squares Fit
krylov: Krylov sequence
linear.pls.fit: Linear Partial Least Squares Fit
normalize: Normalization of vectors
pcr: Principal Components Regression
pcr.cv: Model selection for Princinpal Components regression based on...
pls.cv: Model selection for Partial Least Squares based on...
pls.dof: Computation of the Degrees of Freedom
plsdof-package: Degrees of Freedom and Statistical Inference for Partial...
pls.ic: Model selection for Partial Least Squares based on...
pls.model: Partial Least Squares
ridge.cv: Ridge Regression.
tr: Trace of a matrix
vcov.plsdof: Variance-covariance matrix
vvtz: Projectin operator