This is my general understanding on the affect of Multi-Col linearity on Linear and Logistic regression. However I am still not able to understand whether multi-col linearity is a problem in decision tree (Classification and regression).

Linear Regression - It has a huge impact as the co-efficient might vary drastically if there are 2 variables with high correlation. And regularisation would overcome this issue. Logistic Regression - Same as Linear regression.

Can anybody please give me some insight on how it affects in decision tree classifier and decision tree regressor. And how to overcome this problem if it affects.