Overview and Contributions

Abstract

Outliers are observations that behave in an unusual way with respect to the majority of data, and outlier detection techniques have become an extremely important research branch of modern advanced data mining technologies. Many popular outlier detection algorithms have been developed. The purpose of this book is to introduce some new developments in the unsupervised outlier detection research and some corresponding applications from a k-nearest-neighbor-based perspective. In this chapter, an overview of this book is presented. First, the research issues on unsupervised outlier detection are introduced. Then, the content for each chapter is described. Finally, a summary of our contributions is presented.