I have using rattle on my system and have a couple of queries regarding the same. I am relatively new to rattle and data mining, so I apologize if some of the queries sound too naive but I would greatly appreciate the help. Ok so here I go...
I am not able to use the boost of SVM option on rattle. The moment I execute my data set both of these options become inactive for me.
can I carry out CV in rattle say 5- fold.
I have already split my data in training and test and done variable selection in caret. So now for using RF in rattle, I load and build the model on training set and predict the scores for both training and test to calculate R2 for predicted and observed values. Is that the right way to go about it.
Thank you so much for all your kind suggestions. I look forward to some help on the same.
1) I suspect your dataset may be multi-classification? If so then as I recall, boost and SVM assume (for now) binary classification.
2) Not natively supported - should be but just not implemented yet.
3) If already split, be sure to turn off Rattle's Partition, load your training dataset then in the evaluate tab specify the other dataset you have as the evaluation dataset.