Investigations on big data classification using deep learning

Abstract

The big data generated in healthcare industry is too vast to handle due to which it is challenging to analyse the data and get insights out of it, especially in terms of angiographic illness risk identification. Conventional approaches confront difficulties in the identification of features and low performance, when handling huge datasets. Extreme Learning Machine (ELM) approaches find it too challenging to categorize the big data because of the presence of scattered and continuous blocks. Conventional learning algorithms too lack behind in overcoming the challenges faced in this regard at acceptable speed. The general data mining methodologies are incompetent to handle the issues such as volume, complexity, diversity etc. newlineThese challenges hinder healthcare facilities offered to the end user. While the research community is working upon big data learning to overcome the challenges and develop big data classification algorithms, the current study is aimed at developing a new technique for big data classification named as Rider Chicken Optimization Algorithm-based Recurrent Neural Network (RCOA-based RNN). RCOA-based RNN is an excellent classification approach to classify huge volumes of data on Spark architecture. At first, the data is collected from master node and is disseminated to the slave node. Master and Slave nodes store the data and do the computation process. newline