An Application Approach of Stratified Sampling in Analytic-Predictive Environments of Big Data
7 Pages Posted: 14 Jun 2019
Date Written: March 20, 2019
Abstract
In a simple word, big data is large datasets which are generating from different communication devices such as mobile phone, tablet, laptop, and sensor in the form of structure, unstructured and semi-structured. A few years ago the big data has been generated from social media but now-a-days big data are coming from another sector also like business transactional data from customers and the supply chain. Big data is not in size, it has basically three parameters such as volume, variety and velocity and one more is veracity that is coined by the IBM. Today big data is a complex task to store and process those datasets because the 95% data are unstructured that’s why we need a new tool and techniques for predictive analytics. This paper investigates and to focus is on predictive analytics using stratified random sampling. The predictive analytics basically is based on statistical methods due to huge amounts of data has been stored on different servers so we can obtain small sample from the selected servers and further examine the significance of a particular relationship. The main objective of this research paper to explore the existing sampling techniques which are involved in big data predictive analytics and as well as how to use stratified random sampling in big data predictive analytics.
Suggested Citation: Suggested Citation