R: How to interpret my party plot decision tree?

Question

I made a decision tree with party plot in R but don't really know what it's telling me. The data is about the indian startup ecosystem. I don't know what nodes mean.

Answer 1

The tree displays the subsets found through recursive partitioning. The dependent variable in every resulting subset is visualized by a box plot.

As an example, consider node 10:

• There are 95 observations in this subset.
• These are all the observations which used "seed funding" as the InvestmantnType and had a Date.dd.mm.yyyy in 15, 2018, or 2019.
• The boxplot shows the distribution of the dependent variable Amount.in.USD for these 95 observations.

Answer 2

You can extract the variable importance from tree$variable.importance or you can simply summary(tree) . Variable importance tells us which variables are most important for the models to make decisions.