DATA ANALYSIS AND MACHINE LEARNING EFFORT IN HEALTHCARE

This chapter examines the ideology of data‐driven healthcare and explores how new projects are defined and what is delivered as a proof of concept (POC). It discusses prerequisites and personnel qualifications for data science in healthcare; with a contributed subsection on the role of highperformance computing (HPC). The chapter provides a description of medical data and list the issues arising during data acquisition and transformation; a contributed subsection discusses de‐identification data sharing. It overviews the main themes of machine learning; the intention is to provide the readers with ideas and keywords for future independent exploration. The chapter presents a case study: prediction of rare adverse events based on traditional, nonspecific medical data. The enabling technology is unified electronic medical records (EMRs) that allow access, search, and simple sorting or comparison of hundreds of standard clinical features for millions of patients.