Evolution of Data Analytics in Healthcare

Evolution of Data Analytics in Healthcare

P. Victer Paul (Indian Institute of Information Technology, Kottayam, India), Harika Krishna (Vignan's Foundation for Science, Technology and Research, India) and Jayakumar L. (Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, India)
DOI: 10.4018/978-1-5225-9643-1.ch012

Abstract

In recent years, a huge volume of data has been generated by the sensors, social media, and other sources. Researchers proposed various data analytics models for handling these data and to extract insight that can improve the business of various domains. Data analytics in healthcare (DAiHC) is recent and attracted many researchers due to its importance in improving the value of people's lives. In this perspective, the chapter focuses on the various recent models proposed in DAiHC and dissects the works based on various vital parameters. As an initial part, the work provides comprehensive information on DAiHC and its various application illustrations. Moreover, the study presented in the work categorizes the literature on DAiHC based on factors like algorithms used, application dataset utilized, insight type, and tools used for evaluation of the work. This survey will be helpful for novice to expert researchers who works in DAiHC, and various challenges in DAiHC are also discussed which may help in defining new problems associated with the domain.
Chapter Preview
Top

Introduction

Data analytics are a process of acquiring both subjective, quantifiable models and procedures that can be utilized to improve the business and productivity. Data sets are examined to analyze observable data and patterns in order to produce outcomes as per the organizational needs (DAD, 2019). Data analytics are mainly concentrating and providing solutions for the specific applications like public sector services, healthcare, social networking and education (DAA, 2019). Recent studies anticipated that enterprise server in the world are processing extensive amount of data in the year 2008 (Kambatla, Kollias, Kumar, & Grama, 2014). Our current output of data is approximately 2.5 quintillion bytes of data per day. Almost over 90% of the data has been produced in the world over the last two years (DAR, 2019). Due to advent of Internet of Things (IoT) and ever increasing number of embedded devices, we can estimate that by every two years the data are getting doubled and in coming years the data is going to be unstable. Most of the large volume of the data is generated from health care providers, retail and enterprise.

Analytics is to use the data and find advantageous patterns to make better decisions and it is one of the BI (business intelligence) techniques to improve the organizational gain. Data analytics are the best way for business people grab the customers ‘attention and to get success in business without facing more stress (DAP, 2019). Data analytics models can enhance the businesses by increasing the profits and also operational efficiency. (DAI, 2019). Analytics mainly focused on two vital issues such as the forward motion of the data and value. Currently, data analytics are playing a major role in many organizations by giving useful insights and valid decisions based on their necessities. It is an emerging technique in a recent survey and used as a tool in any industry where data is collected and accessed.

Many more organizations are using data analytics to give qualitative patterns. This data can be used in various circumstances as improving the quality of the organizations, identifying fraud from insights on online and digital information, and improving customer service (AP, 2019). Fig.1 shows the data analytics process, where the knowledge discovery in databases (KDD) is with more clarity. Fayyad and his team members explained the KDD process by performing some operations in which raw data, preprocessing, data cleaning, Exploratory data analysis, models & algorithms, data-driven process, Visualization Graph, decision making (BD,2019). Raw data refer to the primary unstructured data that has been collected based on the use and purpose of the application. Raw data are also known as the source data. The data whatever we collected is unprocessed until it comes in one order. The data coming in one of the manners then only it is easy from the designers. Each of the preprocessing operations performs different tasks in the raw data. It is a method used to eliminate noise from the collected data what we have. We are collecting data from the different locations than it is not realistic to perform analytics (DAI, 2019), (ML, 2019).

Figure 1.

Process flow of data analytics

978-1-5225-9643-1.ch012.f01

Complete Chapter List

Search this Book:
Reset