3
$\begingroup$

I have data on commute times over a specified route over different days during different conditions. Some of the conditions are categorical (e.g., weather, traffic), and some of them are numeric (e.g., time departing from origin). I'd like to find out which of these conditions most strongly correlate with the shortest commute times on any given day of the week (M-F).

Is there a decision tree-like algorithm that would discover the conditions most strongly correlated with the shortest commute times?

$\endgroup$

1 Answer 1

3
$\begingroup$

How about recursive partioning regression trees?

There is a example from R-package rpart, which has numeric target vector with several categorical and numeric IV's.

library(rpart)  
fit = rpart(Price ~ Mileage + Type + Country, data = cu.summary)
$\endgroup$
0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.