Machine Learning & Training

What Is a Train-Test Split?

A train-test split divides a dataset into two parts: one for training the model and another for evaluating it. Keeping the test set separate ensures that performance is measured on data the model has not seen during training. This helps provide an honest estimate of how the model will generalize.