Hi, Does R provide functions for pruning of Additive Regression Tree Models? Suppose I have grown a large set of trees, say 3000 trees, and now want to find a 'computationally more economic' approximation of the function defined by the trees. In the case of a single tree this is pruning. Is there anything similar for (M)ART? In case this matters I use the gbm package for model estimation. Intuitively, at least in the case when the shrinkage parameter is set to a rather small value, there there should be a way to compress the trees since consecutive trees are correlated. Thank you in advance for your help, Vadim P.S. If possible, please reply to me directly since I do not follow the r-help list anymore due to its high traffic [[alternative HTML version deleted]]