Tune multivariate response random forest

Question

Is it possible to tune a random forest ( cforest ) with a multivariate response variable using caret ? eg

mtry_grid <- data.frame(mtry = seq(5,50,5))
train_mtry_class <- train(Class+PRE_POST~., data=rf_data[,-c(1,2)],
                      method='cforest', tuneGrid=mtry_grid, metric='Accuracy')

If not, does anyone have any suggestions for tuning a random forest with a multivariate response?

Answer 1

There's a great CRAN package, which you can do multivariate random forest tuning: https://cran.r-project.org/web/packages/MultivariateRandomForest/MultivariateRandomForest.pdf

Alternatively, you can use "party" (also in CRAN): https://cran.r-project.org/web/packages/party/party.pdf - look at "Conditional Inference Trees"