This book aims to sort out the clear logic of the development of machine
learning-driven privacy preservation in IoTs, including the advantages
and disadvantages, as well as the future directions in this
under-explored domain. In big data era, an increasingly massive volume
of data is generated and transmitted in Internet of Things (IoTs), which
poses great threats to privacy protection. Motivated by this, an
emerging research topic, machine learning-driven privacy preservation,
is fast booming to address various and diverse demands of IoTs. However,
there is no existing literature discussion on this topic in a
systematically manner.
The issues of existing privacy protection methods (differential privacy,
clustering, anonymity, etc.) for IoTs, such as low data utility, high
communication overload, and unbalanced trade-off, are identified to the
necessity of machine learning-driven privacy preservation. Besides, the
leading and emerging attacks pose further threats to privacy protection
in this scenario. To mitigate the negative impact, machine
learning-driven privacy preservation methods for IoTs are discussed in
detail on both the advantages and flaws, which is followed by
potentially promising research directions.
Readers may trace timely contributions on machine learning-driven
privacy preservation in IoTs. The advances cover different applications,
such as cyber-physical systems, fog computing, and location-based
services. This book will be of interest to forthcoming scientists,
policymakers, researchers, and postgraduates.
