Multi-aspect Learning

This book provides a comprehensive review and in-depth discussion on multi-aspect data learning, detailing the state-of-the-art representation learning approaches with a focus on clustering and how these unsupervised approaches are applied to various domains and applications of multi-aspect data. The first time the multi-aspect data is reviewed in a systematic manner where various multi-aspect related concepts and a wide range of applications are fully considered. The first time the application of manifold learning used in dimensionality reduction is investigated thoroughly for multi-view data learning. This book thoroughly presents the state-of-the-art approaches to matrix factorization, subspace clustering, spectral clustering and deep learning methods. These approaches are presented in a manner where the main characteristics and challenges of multi-aspect data are the central focus. Each chapter, in addition to providing state-of-the-art of multi-aspect data learning methods, brings forth a comprehensive discussion of important gaps for future work. Each chapter provides readers with a thorough grasp of the baseline information required for them to apply these methods to future domains and applications as well as innovate novel research in this emerging area.