Early diabetic disease prediction Using mapreduce framework with big

Abstract

Big data analytics has been used for analyzing large volume of data. Big data are significantly employed in healthcare application to determine the normal and abnormal patients. Diabetic disease is a considerable health problem that affects the mortality rate. Hence, early diabetic prediction is essential to provide better treatment. In big data, feature selection has been employed to desire the more relevant features for offering the accurate prediction outcomes. Recently, many research work have been designed for detecting the diabetic diseases. Conventional deep neural network has been developed to determine the normal and abnormal cases of diabetic disease whereas the accuracy has not improved at the required level. Besides, machine learning algorithms with Hadoop-based clusters have been introduced for diabetic prediction. But, the computational time is not reduced. Recurrent convolutional neural network has been introduced to measure the disease risk by gathering the data from the hospitals. However, the performance speedup is not concentrated. Consequently, the novel machine learning techniques are introduced to achieve better diabetic disease prediction. newlineA Normal Discriminant Feature Selection based Regressive Deep Neural MapReduce (NDFS-RDNMR) framework has been proposed and its main objective is to execute diabetic chronic disease prediction with maximum accuracy and precision. The proposed NDFS-RDNMR framework is implemented with the contribution of feature selection or preprocessing and classification processes. Normal Discriminative Preprocessing Model (NDPM) has been employed in NDFS-RDNMR framework to choose more relevant features for disease prediction. In NDPM, the relationship between the relevant features is preserved by means of Min-Max normalization approach via rescaling the features from one range of values to the new range of values and it aids to increases the accuracy while identifying the diabetic chronic disease. newline