Machine Learning Techniques for IoT Data Analytics

Tremendous advancements and innovations in hardware and software together with developments in different communication technologies and computational advancements have encouraged the advent of an ecosystem of highly interconnected and smart devices known as the Internet of Things (IoT). This IoT is getting outstandingly universal and ubiquitous in today's society with a transformative and constructive impact on different application domains like smart and sustainable living, smart healthcare, smart agriculture, smart cities, manufacturing, and more. Consequently, there has been escalating quantitative expansion in the devices connected to IoT leading to the generation of massive volumes of data, depicting a perfect overlap of big data generation with that of IoT. These voluminous data are totally useless without any proper analytic procedure and is much more than humans will be able to process and analyze. In addition to being huge in terms of size and volume, the big data generated by IoT are of highly veracious and variable nature with a variety of data forms and quality. Further, it is highly distinguished by its velocity in terms of creation, time, processing, location dependency, and accessibility. As a result, smart analysis of such data for obtaining valuable insights is a challenging task. Intelligent data analytics is an important prerequisite to maximize the business value of IoT and realize its hyped market potential. Machine learning has the power to handle different challenges associated with IoT data and can be deployed effectively as they require minimal human intervention. In this chapter, different machine learning algorithms will be discussed along with their potential and challenges for IoT data analytics. The central goal of this chapter is to present a taxonomy of various machine learning approaches and demonstrate the implementation of different techniques so as to do higher level analysis using IoT data.