r - how to extract coefficients in logistic regression using gbm? -
i using gbm package generalized boosted regression models, , able extract coefficients produced storage in database.
i using r automatically generate formulas can export database , store. example, have been using dr. harrell's lrm package perform logistic regression, e.g.:
output <- lrm(outcome~predictor1+predictor2,data=dataset) cat(output$coefficients)
is possible gbm? know gbm gives number of trees linearly combined weights there possibility each of tree printed? or perhaps @ least possible in case interaction.depth=1 (e.g., no interactions allowed)?
gbm's (and other tree-based models) don't have coefficients there isn't extract. trying score database using gbm
object? if have 2 options: 1) encode each of gbm
trees sql queries 2) pull data r, score it, , write database.
Comments
Post a Comment