Trees are ubiquitous. No matter what application you are using, chances are that there are trees in play internally. Having said that, not all trees are data structures. In this example, we will explore something different, which is a really popular but not a typical data structure, that is, a decision tree.

At some stage, you would have come across some sort of an automatic prediction system. Whether it is a sports website that predicts the winner of a game or credit scoring websites that tell you which credit card you should apply for to get a quick approval. In this example, we will go with a credit card approval predictor, but this can be easily transferred to any application of your choice.

In a majority of the cases, we have a complex Machine Learning model running behind the scenes to generate an accurate prediction, but, because we know that the number of factors affecting the approval or rejection is finite, we can use a decision tree to determine the chances of approval based on the past patterns exhibited. The following is the list of tasks that we will have to achieve in this example:

1. Creating the decision tree by implementing  Iterative Dichotomiser 3 (ID3) algorithm to classify future samples.
2. Creating a training dataset.
3. Running new inputs through the algorithm and validating responses.