How to choose a split variables for continous features for decision tree

Question

I am currently implementing decision tree algorithm. If I have a continous featured data how do i decide a splitting point. I came across few resources which say to choose mid points between every two points but considering I have 8000 rows of data this would be very time consuming. The output/feature label is having category data. Is any approach where I can perform this operation quicker

Answer 1

Decision tree works calculating entropy and information gain to determine the most important feature. Indeed, 8000 row is not too much for decision tree. But generally, Random forest is similar to decision tree. It is working as ensemble. You can review and try it.Moreover, maybe being slowly is related to another thing.