How to select the optimal split rule? (information gain/entropy)

In order to find the best split rule, we use the information gain, which is derived from the information entropy. This is covered in this video: