A4.3.3 Hyperparameter tuning
Building a classifier is only half the work. The other half is judging whether it is any good and adjusting it until it is as good as it can be. This page covers the four metrics the syllabus names for measuring a classification model, accuracy, precision, recall and F1 score, then explains the two ways a model can fail to generalise, overfitting and underfitting, and finally shows how hyperparameter tuning uses...