is it possible to get the out-of-bag (oob) predictions of each tree, and plot some kind of confidence intervals.
For example this is implemented in the quantregForest
package:
data(ozone,package="gss")
library(quantregForest)
xozone <- ozone[-1]
yozone <- ozone$upo3
qrfozone <- quantregForest(xozone,yozone)
plot(qrfozone)