In previous videos, the dataset used had features which had 3 categories like outlook, temp, humidity. Here the decision tree is a binary tree, which works fine in this case since no feature here has more than 2 categories. So, how do we tackle situations with features having categories more than 2?
Also are the decision trees using sklearn also binary? Or do they have branches at each node according to the number of categories of the feature used in that node?
hey @SanchitSayala ,
Just for a simple understanding about decision trees , the lecture used a binary classes data.
With just the nodes being left or right.
While dealing with various other features with more number of classes. You can use different techniques.
Although there might be any technique , to deal directly with multiple classes feature , but we can easily make it into a binary tree by implementing a two way split on each group of classes.
like for example there 3 classes , then you first split for 2 then again for two to represent 3 way split,
No, its not. There are many different techniques to work . You can search web about them to be more clear.
I hope this had helped you.
Thank You 
.