ABSTRACT
In today's healthcare system, the area of research with the greatest potential is data analysis. The healthcare data are categorized into three different types, namely sensor data, clinical data, and omics data. Clinical data consists of electronic health records, organizational data, claim data, ailment registries, health surveys, and trial data. The sensor data are collected from wearable devices and wireless monitoring devices. The omics data deals with the complete genetic or molecular profiles of human beings. It is not possible to handle sensor, clinical, or omics raw data manually. To handle and analyze the data, machine learning has become a significant tool. In a real-time environment, the data received continuously, from various monitoring devices like life support equipment, ECG, etc., are analyzed using machine learning techniques to predict the results more precisely. Classification is a predictive modeling technique for assigning class labels to the unseen unlabeled samples. Based on the data classification, tasks are assigned into binary classifications, multi-class classifications, multi-label classifications, or imbalanced classifications. A medical diagnostic is a type of imbalanced classification because the data distribution is unequal. This chapter mainly focuses on how classification techniques and predictive tools can be used in the exact prediction of results.
