Smart Healthcare Support Using Data Mining and Machine Learning

Theodora Chatzinikolaou, Eleni Vogiatzi, Anestis Kousis, Christos Tjortjis

Abstract: Ever since the first cities were created, they have been dependent on technology to sustain life. The smart city paradigm integrates advanced monitoring, sensing, communication, and control technologies, aiming at providing real-time, interactive, and intelligent city services to citizens. Thus, the healthcare sector, as an essential part of our lives, could not remain unaffected. The advances in technology provided great potential for many aspects of the health system. In this chapter, we focus on popular machine learning (ML) and data mining (DM) predictive and descriptive techniques and their most prominent applications in the healthcare domain. First, we introduce the process of mining data to extract healthcare relevant knowledge. We briefly review key techniques used, including classification, clustering, and association rule mining. We then focus on specific smart healthcare applications including but not limited to (i) assisting diagnosis and treatment, (ii) health management, (iii) disease prevention and risk monitoring, (iv) virtual assistant and wearable sensors, and (v) drug research. The chapter concludes with a running example of applying well-known ML and DM techniques to a publicly available dataset related to diabetes and a discussion on the impact of DM in healthcare support.