A comprehensive review of an area of machine learning that deals with the use of unlabeled data in classification problems, this text looks at state-of-the-art algorithms, applications benchmark experiments, and directions for future research.