ABSTRACT
This chapter describes theoretical advances of signal processing techniques and showcases applications of both signal processing and machine learning tools for biomedical big data analytics across different domains, including neuroimaging, cardiac, retinal, genomic, sleep, and rehabilitation. Data quality is an often overlooked issue but extremely important in big data analysis, as "bigger data" does not always mean "better data". The health sector is a domain in which data variety plays a crucial challenge for big data analytics. Data can be clinical, genetic, behavioural, environmental, financial, and/or operational. In order to handle such amounts of data, new programming frameworks have been designed to distribute and parallelize data analysis to computing clusters and/or grids. Storage of zettabytes of data is not trivial and requires novel compression and dimensionality reduction methods. Dimensionality reduction is an effective solution to find a meaningful low-dimensional structure from high-dimensional data. The chapter also presents an overview on the key concepts discussed in this book.
