XGBoostParameterGrids
XGBoostParameterGrids(
TaskType = "CPU",
Shuffles = 1L,
NTrees = seq(500L, 5000L, 500L),
Depth = seq(4L, 16L, 2L),
LearningRate = seq(0.05, 0.4, 0.05),
MinChildWeight = seq(1, 10, 1),
SubSample = seq(0.55, 1, 0.05),
ColSampleByTree = seq(0.55, 1, 0.05)
)
"GPU" or "CPU"
The number of shuffles you want to apply to each grid
seq(500L, 5000L, 500L)
seq(4L, 16L, 2L)
seq(0.05,0.40,0.05)
seq(1.0, 10.0, 1.0)
seq(0.55, 1.0, 0.05)
seq(0.55, 1.0, 0.05)
A list containing data.table's with the parameters shuffled and ready to test in the bandit framework