Machine Learning Based big Data Analytics Framework for Healthcare Applications

Abstract

With the expansion of big data (BD) in the health and biomedical sectors, intelligent prediction of medical data promotes early disease identification, patient treatment, and social service. Although when medical information is poor quality or inadequate, analysis accuracy suffers. Also, distinct regional diseases have distinctive traits that vary between regions, which could make it harder to forecast when a disease would spread. In this study, we simplify the machine learning (ML) method termed as a multi-objective firefly optimized improved support vector machine (MFO-ISVM) for accurate chronic illness prediction in places with high disease incidence. We test our suggested strategy using data from actual hospitals. The raw data is first pre-processed using a z-score normalization method. We are aware of no work in the field of healthcare BD analytics that specifically addressed both data types. The suggested MFO-ISVM method has the highest efficiency when compared to other algorithms.