Symmetric Matrix-Based Predictive Classifier for Big Data Computation in Cloud Environment

Abstract

In big data applications, data collection has grown exponentially and it is very complex to extract, identify and transmit information using the existing software tools and it increases the gap in the performance between legitimate classifiers. Cloud computing is a parallel distributed computing system that is frequently used for big data analytics. Data Mining in Big Data improves security and privacy in cloud environment using HACE theorem. Similarly, Flex-Analytics increases the data transmission bandwidth, but contains space and time complexity issues. An efficient PSM-PBC model is proposed for big data computation and information sharing in cloud environment. The sharing of information in the proposed PSM-PBC model includes three processes. Initially, tridiagonal symmetric matrix is constructed in parallel on distributed big data applications that enables faster computation for data extraction and sharing the information across cloud paradigm using Householder transformation. Then Cross-validated bayes classifier model is developed in the proposed PSM-PBC model which evaluates real value diagonal search data for its corresponding query results. Thus, the result obtained from each user request increases the predictive rate. Finally, MapReduce function is enhanced with bayes classes which presents predictive analytics about big data for better data computation and information sharing.