A Study on First Ordered Online Supervised Learning Techniques and Algorithms for Big Data Analytics

Abstract

Big Data Analytics and Machine Learning techniques are emerging streams both in sciences and Industry as well. The characteristics of big data include volume, velocity, visualization, viscosity and variety. Machine learning is at its crux because of its ability to analyse data and provide accurate results. With steady data growth rate, machine learning process can learn more quickly and present promising results. Due to this reason big data analytics and machine learning facilitate each other. Traditional machine learning algorithms like supervised, unsupervised, Reinforcement techniques were not so effective on big data. Map reduce, Batch learning, Online learning, Incremental learning are few of the possible solutions to fit the challenges correlated with big data analytics. The area of online machine learning in big data streams covers algorithms that work on data streams with a limited scope to store past data. In data stream model, past data does not exist to make the heuristic decisions as the fresh data is generated. In this study, an overview of machine learning models for online learning is provided. The most important ideas for classification, regression, recommendation, and supervised modeling from streaming data has been highlighted.