1. Calculate the significance of each feature in data-splitting.
2. Then, pick the feature with the highest significance, and split the data by this. This means that your test will check which of the different values of the feature the example meets. 3. Repeat Step 2 under each node, and so on, until an answer is achieved.
Congratulations, you have just learned the conceptual basis for Decision Trees! In the notebook coming up, we will go through the steps involved in implementing it in Python, but suffice it to say: you have finished the hard part!