All geek questions in one place

The GBM model in h2o produces a different variable value than with R?

When we launched the Gradient Boosting Machine model with Caret package in R, we indicated that some variables (x1, x2, x3) have a higher variable value, but then when we try to run the same GBM in h2o ( http: // h2o2016 .wpengine.com / wp-content / themes / h2o2016 / images / resources / GBMBooklet.pdf ), we get the full set of variables as important. Is there any specific reason for Caret and other factors?

+6
source share

No one has answered this question yet.

See related questions:

Source: https://habr.com/ru/post/1015804/

More articles:


All Articles