I need to do storytelling about which factor drives an athlete to choose full Ironman race. Right now, let's assume Decision Tree performs well so my strategy is to consider the most important feature drives the most.

Since it is known that the important feature of Decision Tree is the one that reduces the impurity of Tree the most, how can I use this theory to explain how much it drives the output? In other words, ideally, I want to have an explanation similar to Linear Regression's (i.e. y = 2$x_1$ + 3$x_2$, that means when $x_2$ is fixed, y will increases 2 units if $x_1$ increases 1).

PS: Here is the data columns and the corresponding explanation, where Resp is the output while others except ID inputs: enter image description here

  • $\begingroup$ My assumption is, a more important feature is more likely to separate the data solely. Looking forward to better explanations. $\endgroup$ Feb 26 at 22:44
  • $\begingroup$ Many more: stats.stackexchange.com/… $\endgroup$
    – Sycorax
    Jul 8 at 14:41