I am working with JMP PRO 2013. I am using partition method( bootstrap forest) for prediction. I also created a validation column which can split my data into training data, test data and validation data. My question is how can i test my data on training data.
For instance in R , you split your data into training and test data and then you predict on the same training data.
It would be great if someone can help me with this.
There are 2 ways to do this. Both are selectable from the primary dialog screen for the Partitioning Platform.
First, you can specify a Validation Portion, and JMP will randomly divide you data into Training and Validation subsets, and run on the Training data and then Validate by running the predictive results on the Validation subset.
The second way, is to specify a "Validation" column by selecting your validation column and placing in the "Validation" role.