Tutorial: build a tiny classifier

In this tutorial you will train a small classifier, evaluate it honestly and explain the result.

1. Frame the problem

Decide what you are predicting and what data you have. Write the goal in one sentence.

2. Split the data

Keep a portion of the data aside as a test set and do not look at it until the end.

3. Start with a baseline

Before anything clever, predict the most common class. Any real model should beat it.

4. Evaluate

Report accuracy and look at the mistakes. A model you understand is worth more than a slightly better one you cannot explain.