Data Mining and Machine Learning Applications

4Stream Mining: Introduction, Tools & Techniques and Applications

Naresh Kumar Nagwani

Computer Science & Engineering, National Institute of Technology Raipur, Raipur, India

Abstract

Stream mining is a key research area in big data analytics, where mining operations are applied on the input streamed data. There are numerous modern applications of it, particularly after the invention of the Internet of Things (IoT), which generates the real time streamed data from various sources. Storage of the streamed input data in a local system is not feasible because it enters in continuous fashion and volume of the data is very high, so an analysis of this data requires the single pass processing. Data selection and summarization are the major challenges in mining streamed data. This chapter provides an overview of stream mining and provides a brief introduction of various tools and techniques available for implementing mining operations on streamed data. Major stream mining techniques such as sampling and sketching from data streams, concept drift detection, classification, clustering, frequent set mining and outlier detection techniques are discussed in this chapter. A, overview of tools for processing the data stream in Java, Python and R programming languages is also presented. Some of the key application areas of stream mining are also discussed.

Keywords: Stream mining, stream clustering, stream implementation, applications of stream mining, tools and technologies for stream mining ...