Skip to contents

Details

Four possible SVMs can be implemented, dependent on the type parameter. These correspond to predicting the survival time via regression (regression), predicting a continuous rank (vanbelle1, vanbelle2), or a hybrid of the two (hybrid). Whichever type is chosen determines how the crank predict type is calculated, but in any case all can be considered a valid continuous ranking.

makediff3 is recommended when using type = "hybrid".

Dictionary

This Learner can be instantiated via the dictionary mlr_learners or with the associated sugar function lrn():

mlr_learners$get("surv.svm")
lrn("surv.svm")

Meta Information

  • Task type: “surv”

  • Predict Types: “crank”, “response”

  • Feature Types: “integer”, “numeric”, “character”, “factor”, “logical”

  • Required Packages: mlr3extralearners, survivalsvm

Parameters

IdTypeDefaultLevelsRange
typecharacterregressionregression, vanbelle1, vanbelle2, hybrid\((-\infty, \infty)\)
diff.methcharacter-makediff1, makediff2, makediff3\((-\infty, \infty)\)
gamma.mulist-\((-\infty, \infty)\)
opt.methcharacterquadprogquadprog, ipop\((-\infty, \infty)\)
kernelcharacterlin_kernellin_kernel, add_kernel, rbf_kernel, poly_kernel\((-\infty, \infty)\)
kernel.parslist-\((-\infty, \infty)\)
sgf.svinteger5\([0, \infty)\)
sigfinteger7\([0, \infty)\)
maxiterinteger20\([0, \infty)\)
marginnumeric0.05\([0, \infty)\)
boundnumeric10\([0, \infty)\)
eig.tolnumeric1e-06\([0, \infty)\)
conv.tolnumeric1e-07\([0, \infty)\)
posd.tolnumeric1e-08\([0, \infty)\)

References

Belle VV, Pelckmans K, Huffel SV, Suykens JAK (2010). “Improved performance on high-dimensional survival data by application of Survival-SVM.” Bioinformatics, 27(1), 87–94. doi: 10.1093/bioinformatics/btq617.

Belle VV, Pelckmans K, Huffel SV, Suykens JA (2011). “Support vector methods for survival analysis: a comparison between ranking and regression approaches." Artificial Intelligence in Medicine, 53(2), 107–118. doi: 10.1016/j.artmed.2011.06.006.

Shivaswamy, P. K., Chu, W., & Jansche, M. (2007). A support vector approach to censored targets. In Proceedings - IEEE International Conference on Data Mining, ICDM (pp. 655–660). https://doi.org/10.1109/ICDM.2007.93

See also

Author

RaphaelS1

Super classes

mlr3::Learner -> mlr3proba::LearnerSurv -> LearnerSurvSVM

Methods

Inherited methods


Method new()

Creates a new instance of this R6 class.

Usage


Method clone()

The objects of this class are cloneable with this method.

Usage

LearnerSurvSVM$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

if (requireNamespace("survivalsvm", quietly = TRUE)) {
  learner = mlr3::lrn("surv.svm")
  print(learner)

  # available parameters:
  learner$param_set$ids()
}
#> <LearnerSurvSVM:surv.svm>
#> * Model: -
#> * Parameters: list()
#> * Packages: mlr3, mlr3extralearners, survivalsvm
#> * Predict Type: crank
#> * Feature types: integer, numeric, character, factor, logical
#> * Properties: -
#>  [1] "type"        "diff.meth"   "gamma.mu"    "opt.meth"    "kernel"     
#>  [6] "kernel.pars" "sgf.sv"      "sigf"        "maxiter"     "margin"     
#> [11] "bound"       "eig.tol"     "conv.tol"    "posd.tol"