IoT Architecture for End‐to‐End Visibility

Abstract

This chapter provides a methodical basis of how predictive analytics works in theory and practice around the field related to consumer electronics. It also commences with an overview of the fundamental concepts, namely, data mining, machine learning, and statistical modeling, and the way they contribute to the success in forecasting. Then, the categories of the predictive models (classification, regression, which is clustering) are contrasted with the case of their use. Then, it is going to talk about stuff such as source data, such as usage logs, normalized behavioral analytics feeds from real-time sensor feeds into device system. The chapter highlights some of the real practical uses, such as predicting consumer preferences, enriching the features of products depending on their use patterns, incorporating Internet of Things–based diagnostic capabilities, producing predictive capabilities, predictive demand, and personalization. Last but not least, the examples of case studies within the industry and the analysis of the development of the artificial intelligence algorithms, e.g., deep learning, federated learning, and edge computing, are given to demonstrate the current trends. Finally, the role of predictive analytics as one of the crucial technologies that can enhance the performance of design decisions, product life cycle, user experience, and further contribution to sustainability and resource efficiency is introduced.