Supervised learning

Classification is the assignment of objects (data points) to categories (classes). Regression allows you to assign a continuous output to your data by estimating the relationship between your features and the output.
Both require a training data-set of points with known class labels and a test data-set for evaluation.