R GBM：从二项结果的网格搜索中检索最优模型（伯努利分布）

我在R中使用GBM3包来提升梯度。我已经建立了一个网格搜索模型。现在，我正在尝试获得最佳模型，但不知道如何检索与贝努利分布相关的指标，如AUC或logLoss。有人能帮我解决这个问题吗

# Create grid search
 hyper_grid <- expand.grid(shrinkage = c(0.3, 0.1, 0.05, 0.01, 0.001),interaction.depth = c(3,5),n.minobsinnode = c(10, 15),bag.fraction = c(.65,0.5, .8, 1),trees = NA,time = NA)

# Total number of combinations
 nrow(hyper_grid)

# Train model
for(i in 1:nrow(hyper_grid)) {
set.seed(123)
gbm.tune <- gbm(
formula = result ~ .,
distribution = "Bernoulli",
data = Train,
n.trees = 10000,
interaction.depth = hyper_grid$interaction.depth[i],
shrinkage = hyper_grid$shrinkage[i],
n.minobsinnode = hyper_grid$n.minobsinnode[i],
bag.fraction = hyper_grid$bag.fraction[i],
train.fraction = .75,
verbose = FALSE)

hyper_grid$optimal_trees[i] <- which.min(gbm.tune$valid.error)}


arrange(hyper_grid, metric???)

#创建网格搜索
超网格