Publication Details
Abstract
Big data is a huge amount of data that is such a large amount that it is difficult to process using conventional methods of database and software. When using big data-related applications technical barriers are encountered when moving data between different locations that is costly and requires massive main memory for processing. Big data is a term used to describe interactions and transactions of data in relation to their magnitude and complexity that go beyond the normal technical capability of the capture, organization and processing of data within the cloud. It features real-time processing of data which runs in high-performance clusters. Applications that use big data are designed to share structured and unstructured information. They collect the data in a way that allows for speedier response and reduce the time for classification. Similarly, in this paper, a Discretized Support Vector Classification and Prediction (EEDSV-CP) model is suggested to provide effectual computation upon huge data apps and sharing in a cloud computing environment. Originally, pre-processing was carried out in the EEDSV-CP model using interval equivalence discretization, which aids in the removal of noise and erratic data obtained out of various sources. The computation temporal and spatial complexity are mitigated out by denoising and inconsistizing the data. Furthermore, the EEDSV-CP model employs a supportive vector prediction classifier to categorize data centered upon user query request by employing parallel hyperplanes, with the aim of increasing classification accuracy of customer data requesting on big data. The proposed EEDSV-CP precisely predicts the customer data requesting on big data with the classified data.