Decision Trees and Overfitting

Decision trees are one of the most beginner-friendly models in machine learning—they mirror how we make choices in everyday life: step-by-step, rule-by-rule. That’s part of why they are commonly used in interviews. In this lesson, we’ll walk through building a simple decision tree to classify sentiment in text, using basic word counts and logic. Along the way, we’ll see how this approachable model can easily overfit and how to keep it grounded. Let's begin!

Decision trees for sentiment analysis

Our company specializes in analyzing customer reviews to improve product feedback loops. We’re hiring data scientists and machine learning engineers who can design simple, interpretable models that perform well under constraints.

You’ve been asked to create a lightweight sentiment classifier that uses decision trees to predict whether a review is positive, negative, or neutral based on specific keywords. ...