TentativeRoughFix: Rough fix of Tentative attributes

In Boruta: Wrapper Algorithm for All Relevant Feature Selection

Rough fix of Tentative attributes

Description

In some circumstances (too short Boruta run, unfortunate mixing of shadow attributes, tricky dataset...), Boruta can leave some attributes Tentative. TentativeRoughFix performs a simplified, weaker test for judging such attributes.

Usage

TentativeRoughFix(x, averageOver = Inf)

Arguments

x	an object of a class Boruta.
averageOver	Either number of last importance source runs to average over or Inf for averaging over the whole Boruta run.

Details

Function claims as Confirmed those attributes that have median importance higher than the median importance of maximal shadow attribute, and the rest as Rejected. Depending of the user choice, medians for the test are count over last round, all rounds or N last importance source runs.

Value

A Boruta class object with modified finalDecision element. Such object has few additional elements:

originalDecision	Original finalDecision.
averageOver	Copy of averageOver parameter.

Note

This function should be used only when strict decision is highly desired, because this test is much weaker than Boruta and can lower the confidence of the final result.

x has to be made with holdHistory set to TRUE for this code to run.

Boruta documentation built on Nov. 12, 2022, 9:06 a.m.