In supervised learning, the algorithm is developed using data that contain a label (dependent variable or event) and independent features (variables). The algorithm then predicts future or unknown values of the label of interest, using features (independent variables).
For instance, a data set of counterparties may contain labels on some data points identifying those that are in default and those that are not in default. The algorithm will learn a general rule of classification that it will use to predict the labels for other observations in the data set. Some of the supervised techniques include regression, decision trees, random forests, gradient boosting and deep neural networks.