I am using gbm for one of the classification problems and my code to predict the responses is:
pred.gbm <- predict(fit.gbm, newdata=test.rose[,1:17], n.trees=best.iter, type="response") pred.label <- 1*(pred.gbm > 0.484) #1: > 0.484; 0: otherwise
The last line is something i wish to implement but how do I find out the cutoff value,like from a specificity vs sensitivity curve?
head(pred.gbm)  1.965139 1.123409 1.277245 1.001579 1.624551 1.834146
I want the predicted probabilities rather than these values.Can the reason be that I am using
distribution = "laplace" and not “bernoulli”??
Can someone please help me with these??