Supervised Learning

Supervised learning is the most widely used machine learning paradigm. Models are trained on datasets of input-output pairs, where each input is associated with a known correct label or target value.

The model learns a mapping by minimizing a loss that measures the gap between predictions and ground truth.

Many core tasks use this paradigm, including classification, regression, and object detection. Results depend strongly on label accuracy and dataset size.

Trained models must be validated on a separate test set to confirm their ability to generalize to new samples.