Dimensionality reduction (DR) refers to the problem of projecting high-dimensional data onto a low-dimensional manifold so that relevant information is preserved. DR arises in many application areas where direct processing of the data is too costly. Through a machine-learning perspective that focuses on algorithms rather than theory, Dimensionality Reduction provides an overview of methods for DR including real-world applications taken from areas such as speech processing and computer vision. Interest in this area has exploded in recent years, making it a growing field of research. This book serves as the first reference for interested graduate students and researchers.