A Novel Multi-Objective Learning Algorithm for Disease Identification and Classification in Electronic Healthcare System

Abstract

Data is a commodity in today’s electronic world, and massive amount of data is being generated in many fields. Medical files and disease-related data are two types of data in the healthcare industry. This electronics health data and machine learning methods would enable us all to evaluate vast amount of data in order to uncover hidden patterns in disease, offer individualized treatment to the patients, and anticipate disease progression. In this paper, a general architecture for illness prediction in the health industry is proposed. The Internet of Things (IoT), as a helpful model wherein reduced electronics body sensors and smart multimedia medical equipment, are used to enable remote monitoring of body function, plays a critical role, particularly in areas when medical care centers are few. To tackle these challenges, we have proposed Deep Reinforcement Learning with Gradient-based Optimization (DRL with BRO) model for various disease detection and classification such as skin disease, lung disease, heart, and liver disease. Initially, the IoT-enabled data are collected and stored in the cloud storage. After that, the medical decision support system based DRL with the GBO model classifies various diseases. The maximum classification accuracy with the minimum delay is the multi-objective function and finally, the proposed study satisfies the objective functions. Based on the experimental study, the proposed method offers good results than other existing methods.