Parameters for Resubstitution Error Calculation

Share:

Description

Some feature selection functions provided in the framework use resubstitution error rate to choose the best number of features for classification. This class stores parameters related to that process

Constructor

ResubstituteParams() Creates a default ResubstituteParams object. The number of features tried is 100, 200, 300, 400, 500. The performance measure used is the balanced error rate.

ResubstituteParams(nFeatures, performanceType, better = c("lower", "higher")) Creates a ResubstituteParams object, storing information about the number of top features to calculate the performance measure for, the performance measure to use, and if higher or lower values of the measure are better.

nFeatures

A vector for the top number of features to test the resubstitution error for.

performanceType

Either "balanced" or one of the options provided by performance.

better

Either "lower" or "higher". Determines whether higher or lower values of the performance measure are desirable.

intermediate

Character vector. Names of any variables created in prior stages by runTest that need to be passed to classifier.

...

Other named parameters which will be used by the classifier.

Author(s)

Dario Strbenac

Examples

1
  ResubstituteParams(nFeatures = seq(25, 1000, 25), performanceType = "err", better = "lower")

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.