Homomorphic Encryption for Data Science (HE4DS)

This book includes 4 parts. The first part provides the basic knowledge that an application developer should be aware of, in order to understand the Fully Homomorphic Encryption FHE technology. The authors present various techniques to leverage the unique features of FHE and to overcome its characteristic limitations in the second and third parts of this book. Specifically, the second part summarizes in one section many polynomial approximation techniques that are often used by FHE applications, and the third part discusses some packing schemes based on a data structure called tile tensors. Finally, part four demonstrates how to use the studied techniques in several specific privacy preserving inference applications. Examples and exercises are included throughout this book. The proliferation of practical FHE technology has triggered a wide interest in the field and a common wish to experience and understand it. Still, we see that most books and scientific papers related to FHE target a cryptographic audience and thus make it hard for lay readers or readers from other fields of expertise to enter the domain. This book aims to simplify the FHE world for those who are interested in privacy preserving data science tasks. This book is designed as a textbook mainly for under-graduate and graduate-level students in computer science (ML or DS). Data scientists who plan to work on secure/encrypted models will want to purchase this book as a reference as well as applied cryptographer, software/hardware engineers and CISO / CPO / Cyber Security & Privacy SMEs