Healthcare analytics: overcoming the barriers to health information using machine learning algorithms

Abstract

In recent years, the data processed by the healthcare domain is increasing at an unprecedented rate accompanied by rich knowledge for medical research but low information has led to Healthcare Analytics. Healthcare Analytics collects the data from a myriad of areas such as clinicians, hospitals, government agencies, health insurance, pharmaceutical, and biotechnology agencies and allows for the examination of trends and patterns in various healthcare data. Based on this pattern, healthcare analytics determines how healthcare can be upgraded while constraining exorbitant spending. The retrieval of a pattern from healthcare data pertinent to the healthcare application and input to the Machine Learning algorithm is decided based on the data features. Feature selection approaches are developed to choose a subset of features that explain the details to achieve a more appropriate and compact depiction of the knowledge available. This paper summarizes Feature Selection algorithms and presents the challenges involved in healthcare data and also present an abstract architecture of data analytics in healthcare domain. Various algorithms and techniques in machine learning are compared and classified based on the applications of Healthcare Analytics.