In this Chapter the motivation for the setting of adaptive many-core machines able to deal with big machine learning challenges is emphasized. A framework for inference in Big Data from real-time sources is presented as well as the reasons for developing high-throughput Machine Learning (ML) implementations. The chapter gives an overview of the research covered in the book spanning the topics of advanced ML methodologies, the GPU framework and a practical application perspective. The chapter describes the main Machine Learning (ML) paradigms, and formalizes the supervised and unsupervised ML problems along with the notation used throughout the book. Great relevance has been rightfully given to the learning problem setting bringing to solutions that need to be consistent, well-posed and robust. In the final of the chapter an approach to combine supervised and unsupervised models is given which can impart in better adaptive models in many applications.
CITATION STYLE
Lopes, N., & Ribeiro, B. (2015). Motivation and Preliminaries. In Studies in Big Data (Vol. 7, pp. 3–13). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-319-06938-8_1
Mendeley helps you to discover research relevant for your work.