2. Supervised learning

Supervised learning [H 5 010] is a branch of machine learning in which a model is trained on "labeled" data, i.e. examples of inputs associated with known desired outputs. The aim is to teach the model to generalize to new, unlabeled data and make accurate predictions. Supervised learning techniques therefore require prior data processing: the quality and relevance of predictions made using supervised learning algorithms depend on this processing task, which involves ensuring the integrity, relevance and accuracy of the data used.