RunGroupCrossvalidation: Cross-validation wrapper for 'TreeGuidedGroupLasso'.

Description Usage Arguments Value See Also

Description

Perform k-fold cross-validation to find optimal parameters settings.

Usage

1
2
3
4
RunGroupCrossvalidation(X = NULL, task.specific.features = list(), Y,
  groups, weights.matrix, lambda.vec, num.folds = 10, num.threads = 1,
  verbose = -1, row.weights = NULL, standardize = TRUE,
  fit.intercept = TRUE, ...)

Arguments

X

Column centered N by J input matrix of features common to all tasks.

task.specific.features

List of features which are specific to each task. Each entry contains an N by J2 column-centered matrix for one particular task (where columns are features). List has to be ordered according to the columns of Y.

Y

Column centered N by K output matrix for every task.

groups

V by K matrix determining group membership: Task k in group v iff groups[v,k] == 1 weights.matrix: Numerical matrix where each row is a numerical vector mapping groups to their weight.

weights.matrix

Numerical matrix where each row is a numerical vector mapping groups to their weight.

lambda.vec

Vector of regularization parameters.

num.folds

Number of folds.

num.threads

Number of threads to use.

verbose

(Optional) Integer in -2, -1, 0,1,2. verbose = -2: No output. verbose = -1: Display total elapsed time. verbose = 0: Display elapsed time for every parameter. verbose = 1: Print summary at the end of the optimization. verbose = 2: Print progress during optimization.

row.weights

(Optional) Use weighted MSE.

standardize

(Optional) Default is TRUE. Standardize data (using R function scale()). Coefficients will be returned on original scale.

fit.intercept

(Optional) Default is TRUE. Include intercept.

...

Additional parameters passed to TreeGuidedGroupLasso.

Value

List containing

cv.results

data.frame with cross-validation errors for different parameters.

full.model

Full model trained on the whole data set.

See Also

TreeGuidedGroupLasso


tohein/linearMTL documentation built on May 17, 2019, 8:22 a.m.