Use to build a partition-based model (Decision Tree) that identify the most important factors that predict a continuous outcome and use the resulting tree to make prediction for new observations.
Regression Trees
- From an open JMP® table, select Analyze > Predictive Modeling > Partition.
- Select a continuous response variable from Select Columns and click Y, Response.
- Select explanatory variables and click X, Factor.
- If desired, enter the Validation Portion or select a validation column and click Validation (JMP Pro only). A validation set was not used in this illustration.
- In JMP Pro only, select the tree Method: Decision Tree (the default), Bootstrap Forest, Boosted Tree, K Nearest Neighbors, or Naive Bayes.
- Click OK. JMP initially displays a graph of showing the response for all the observations and a line drawn at the overall mean ($3,971 in this example).
- Click the Split button. The original observations will be split into two nodes, or leaves. In the top graph, horizontal lines are drawn at the mean response within each leaf and vertical lines depict the leaf’s relative size.
- Click Split to make an additional split. Click Prune to remove a split. If a validation portion or validation column is used, click Go to perform automatic splitting and pruning optimizing the fit on the validation data.
Notes:
For additional options, such as Leaf Report, Small Tree View, Column Contributions, click the top red triangle. Other options, such as Save Prediction formula and Make SAS® DATA Step, are available from the top red triangle > Save Columns. For split options for a particular node, click on the red triangle for that node.
Interpretation for the first two splits (Response is Diamond Price in $):
• There are 1,502 obs with Carat Weight < 0.95. The mean price of these obs is $2,281.
• There are 1,188 obs with Carat Weight >= 0.95. The mean price of these obs is $6,109.
• For the 1,502 obs in with Carat Weight <0.95, the second split, also based on the Carat Weight variable, is at a Carat Weight of 0.74.
Visit Predictive and Specialized Models > Partition Models in JMP Help to learn more.