Abstract
The CKKS homomorphic cipher (Cheon, Kim, Kim and Song, Asiacrypt 2017) enables computation on vectors of high-dimensional complex numbers, without disclosing them. Its elementary operations are coordinate-by-coordinate addition, multiplication and conjugation, as well as cyclic rotation of coordinates.
We'll show you how to perform these operations, then use them to evaluate polynomials and multiply a matrix by a vector. With these tools at our disposal, we will describe CKKS's "bootstrap", which enables arbitrary calculations to be performed. Finally, we'll look at applications of CKKS, notably in biometrics and machine learning.
