These parameters are auxiliary to tree-based models that use the "C5.0"
engine. They correspond to tuning parameters that would be specified using
set_engine("C5.0", ...)
.
confidence_factor(range = c(-1, 0), trans = log10_trans()) no_global_pruning(values = c(TRUE, FALSE)) predictor_winnowing(values = c(TRUE, FALSE)) fuzzy_thresholding(values = c(TRUE, FALSE)) rule_bands(range = c(2L, 500L), trans = NULL)
range | A two-element vector holding the defaults for the smallest and largest possible values, respectively. |
---|---|
trans | A |
values | For |
To use these, check ?C50::C5.0Control
to see how they are used.
confidence_factor()#> Confidence Factor for Splitting (quantitative) #> Transformer: log-10 #> Range (transformed scale): [-1, 0]no_global_pruning()#> Skip Global Pruning? (qualitative) #> 2 possible value include: #> TRUE and FALSEpredictor_winnowing()#> Use Initial Feature Selection? (qualitative) #> 2 possible value include: #> TRUE and FALSEfuzzy_thresholding()#> Use Fuzzy Thresholding? (qualitative) #> 2 possible value include: #> TRUE and FALSErule_bands()#> Number of Rule Bands (quantitative) #> Range: [2, 500]