Contains linear and nonlinear regression methods based on Partial Least Squares and Penalization Techniques. Model parameters are selected via cross-validation, and confidence intervals ans tests for the regression coefficients can be conducted via jackknifing.
copied from cf-post-staging / r-ppls