In this post, we aim to predict whether a passenger survived the Titanic shipwreck. But rather than training a sophisticated model, we don’t even look at the data. This way, we discover how easy it is to produce seemingly good results that are, in fact, worthless.

Predicting survival on board the Titanic is a classification task since there are distinct outcome values. Specifically, it is a binary classification task because there are exactly two possible predictions (survived or died).

Before we develop a sophisticated machine learning algorithm, let’s implement the simplest algorithm we can imagine. A classifier that guesses.

#machine-learning #programming #data-science #evaluation-metric #python