Model based recursive partitioning - randomized subset of partition variables considered during each split.
The mob function in party package is modified so that a random subset of predictor variables are considered during each split. mtry represents the number of predictor variables to be considered during each split.
mob.rf.tree(main_model, partition_vars, mtry, weights, data = list(), na.action = na.omit, model = glinearModel, control = mob_control(), ...)
A model in character format
A vector of partition variables
A Random subset of partition variables to be considered at each node of decision tree
An optional vector of weights, as described in mob
A data frame containing the variables in the model.
A function which indicates what should happen when the data contain NAs, as described in mob
A list with control parameters as returned by mob_control
Additional arguments passed to the fit call for the model.
Achim Zeileis, Torsten Hothorn, and Kurt Hornik (2008). Model-Based Recursive Partitioning. Journal of Computational and Graphical Statistics, 17(2), 492-514.