martCV {mseq} | R Documentation |
Get the cross-validation R squared for MART.
martCV(data, fold = 5, seed = 281142, shrinkage = 0.06, interaction.depth = 10, n.trees = 2000, small_count = 0.5)
data |
the data frame generated by expData.R |
fold |
number of fold for CV |
seed |
seed to generate training indexes |
shrinkage |
the shrinkage argument of gbm |
interaction.depth |
the interaction.depth argument of gbm |
n.trees |
the n.trees argument of gbm |
small_count |
the small count which will replace the zero count |
the CV R squared, a numeric value
# read and expand the data data(g1_part) # for real data, please use read.csv, like g1 <- read.csv("g1.csv") data <- expData(g1_part, 2, 3) # In real datasets, surrounding sequences should be set longer. # To shorten the running time, this example uses small values of interaction.depth and n.trees. For real datasets, it is strongly suggested to use their default values. # get the CV R squared for MART R_sq <- martCV(data, interaction.depth = 2, n.trees = 100)