mlr_tuning_spaces_rbv1: RandomBot Tuning Spaces

In mlr3tuningspaces: Search Spaces for 'mlr3'

RandomBot Tuning Spaces

Description

Tuning spaces from the Kuehn (2018) article. The hyperparameter respect.unordered.factors and min.node.size of the ranger tuning space differ from the paper.

glmnet tuning space

• alpha [0, 1]

• s [1e-04, 1000] Logscale

kknn tuning space

• k [1, 30]

ranger tuning space

• num.trees [1, 2000]

• replace [TRUE,FALSE]

• sample.fraction [0.1, 1]

• mtry.ratio [0, 1]

• respect.unordered.factors [“ignore”, “order”, “partition”]

• min.node.size [1, 100]

mtry.power is replaced by mtry.ratio.

rpart tuning space

• cp [0, 1]

• maxdepth [1, 30]

• minbucket [1, 60]

• minsplit [1, 60]

svm tuning space

• kernel [“linear”, “polynomial”, “radial”]

• cost [1e-04, 1000] Logscale

• gamma [1e-04, 1000] Logscale

• degree [2, 5]

xgboost tuning space

• nrounds [1, 5000]

• eta [1e-04, 1] Logscale

• subsample [0, 1]

• booster [“gblinear”, “gbtree”, “dart”]

• max_depth [1, 15]

• min_child_weight [1, 100] Logscale

• colsample_bytree [0, 1]

• colsample_bylevel [0, 1]

• lambda [1e-04, 1000] Logscale

• alpha [1e-04, 1000] Logscale

Source

Kuehn D, Probst P, Thomas J, Bischl B (2018). “Automatic Exploration of Machine Learning Experiments on OpenML.” 1806.10961, https://arxiv.org/abs/1806.10961.

mlr3tuningspaces documentation built on April 20, 2023, 5:07 p.m.