Geek Answers Handbook

"real deviation" is for the GBM model, what does it mean and how to get rid of it?

I use Gradient boosting for classification. Although the result is improving, but I get NaN in validdeviance.

Model = gbm.fit( x= x_Train , y = y_Train , distribution = "bernoulli", n.trees = GBM_NTREES , shrinkage = GBM_SHRINKAGE , interaction.depth = GBM_DEPTH , n.minobsinnode = GBM_MINOBS , verbose = TRUE )

Result

enter image description here

How to configure a parameter to get a validdeviance.

+5
source share
1 answer

I had the same problem, oddly enough, we are a little on this ...

Adding train.fraction = 0.5 to the list of options solves the problem (it seems that there is no default value, and validdeviance is not calculated without the train.fraction value explicitly mentioned).

+5
source

More articles:


All Articles