Classification is the most common task in supervised learning paradigm. In this chapter, we first present the problems and definition, and the working principle using formal and illustrated descriptions. We second discuss the basic types including one-class, two-class, multi-class, multi-label, and imbalanced classifications and related elements such as linear and nonlinear, hard and soft, as well as lazy and eager classifications. Next, we focus on several typical classification algorithms including logistic regression, naive Bayes, AdaBoost, support vector machine, and artificial neural network. We then introduce the loss functions covering 0–1 loss, logistic loss, exponential loss, hinge loss, and cross-entropy loss. After that, we discuss some evaluation metrics such as accuracy, precision and recall, sensitivity and specificity, F1 score, precision-recall curve (PR curve), receiver operating characteristic curve (ROC curve), and Matthew’s correlation coefficient (MCC). Finally, we introduce the application fields of classification briefly.