Bagged Logic Regression
method = 'logicBag'
Type: Regression, Classification
Tuning Parameters: nleaves (Maximum Number of Leaves), ntrees (Number of Trees)
Boosted Linear Model
method = 'bstLs'
Type: Regression, Classification
Tuning Parameters: mstop (# Boosting Iterations), nu (Shrinkage)
Cubist
method = 'cubist'
Type: Regression
Tuning Parameters: committees (#Committees), neighbors (#Instances)
Elasticnet
method = 'enet'
Type: Regression
Tuning Parameters: fraction (Fraction of Full Solution), lambda (Weight Decay)
glmnet
method = 'glmnet'
Type: Regression, Classification
Tuning Parameters: alpha (Mixing Percentage), lambda (Regularization Parameter)
Independent Component Regression
method = 'icr'
Type: Regression
Tuning Parameters: n.comp (#Components)
Least Angle Regression
method = 'lars'
Type: Regression
Tuning Parameters: fraction (Fraction)
Least Angle Regression
method = 'lars2'
Type: Regression
Tuning Parameters: step (#Steps)
Linear Regression
method = 'lm'
Type: Regression
No Tuning Parameters
Linear Regression with Backwards Selection
method = 'leapBackward'
Type: Regression
Tuning Parameters: nvmax (Maximum Number of Predictors)
Linear Regression with Forward Selection
method = 'leapForward'
Type: Regression
Tuning Parameters: nvmax (Maximum Number of Predictors)
Linear Regression with Stepwise Selection
method = 'leapSeq'
Type: Regression
Tuning Parameters: nvmax (Maximum Number of Predictors)
Linear Regression with Stepwise Selection
method = 'lmStepAIC'
Type: Regression
No Tuning Parameters
Logic Regression
method = 'logreg'
Type: Regression, Classification
Tuning Parameters: treesize (Maximum Number of Leaves), ntrees (Number of Trees)
Model Rules
method = 'M5Rules'
Type: Regression
Tuning Parameters: pruned (Pruned), smoothed (Smoothed)
Model Tree
method = 'M5'
Type: Regression
Tuning Parameters: pruned (Pruned), smoothed (Smoothed), rules (Rules)
Partial Least Squares
method = 'kernelpls'
Type: Regression, Classification
Tuning Parameters: ncomp (#Components)
Partial Least Squares
method = 'pls'
Type: Regression, Classification
Tuning Parameters: ncomp (#Components)
Partial Least Squares
method = 'simpls'
Type: Regression, Classification
Tuning Parameters: ncomp (#Components)
Partial Least Squares
method = 'widekernelpls'
Type: Regression, Classification
Tuning Parameters: ncomp (#Components)
Penalized Linear Regression
method = 'penalized'
Type: Regression
Tuning Parameters: lambda1 (L1 Penalty), lambda2 (L2 Penalty)
Principal Component Analysis
method = 'pcr'
Type: Regression
Tuning Parameters: ncomp (#Components)
Relaxed Lasso
method = 'relaxo'
Type: Regression
Tuning Parameters: lambda (Penalty Parameter), phi (Relaxation Parameter)
Relevance Vector Machines with Linear Kernel
method = 'rvmLinear'
Type: Regression
No Tuning Parameters
Ridge Regression
method = 'ridge'
Type: Regression
Tuning Parameters: lambda (Weight Decay)
Ridge Regression with Variable Selection
method = 'foba'
Type: Regression
Tuning Parameters: k (#Variables Retained), lambda (L2 Penalty)
Robust Linear Model
method = 'rlm'
Type: Regression
No Tuning Parameters
Sparse Partial Least Squares
method = 'spls'
Type: Regression, Classification
Tuning Parameters: K (#Components), eta (Threshold), kappa (Kappa)
Supervised Principal Component Analysis
method = 'superpc'
Type: Regression
Tuning Parameters: threshold (Threshold), n.components (#Components)
Support Vector Machines with Linear Kernel
method = 'svmLinear'
Type: Regression, Classification
Tuning Parameters: C (Cost)
The lasso
method = 'lasso'
Type: Regression
Tuning Parameters: fraction (Fraction of Full Solution)