A Novel Big Data Approach Using Fuzzy Rule Based Multilayer Perceptrons

Abstract

They are faced with immense quantities and high velocity of data with complicated structures in the big data era. Social networks, sensors, online and offline transactions, and our daily lives can all produce data. When big data is processed correctly, it can lead to relevant, helpful and useful decisions being made in a number of areas, including government, business, management, and medicine and healthcare. Large amounts of data on healthcare have the ability to significantly enhance patient outcomes, predict epidemics, provide insightful information, prevent diseases that may be prevented, reduce the cost of healthcare delivery, and generally increase life. Big data is made up of patient data that is gathered for remote healthcare applications that differs in terms of volume, velocity, variety, veracity, and value.  Healthcare data classification presents a number of challenges for big data since it gathers huge quantities of data. Processing a heterogeneous collection of this size requires a specialized approach, making it one of the most difficult challenges. The paper presents a novel big data approach using fuzzy rule-based multilayer perceptrons to address these problems. Big data offers the ability to accumulate, analyze, manage, and integrate large amounts of disparate, structured, and unstructured information generated by the healthcare systems of currently. A FRCNN (Fuzzy Region based Convolutional Neural Network) classifier is designed to perform normal and disease classification. Accuracy, precision, recall, and F1-score are only some of the performance criteria used to evaluate this model.