Data Science Fellows Projects 2019

Machine learning algorithm adaptation for homomorphically-encrypted data and models

Project by Roey

Abstract

Machine learning algorithm adaptation for homomorphically-encrypted data and models: implement and compare machine learning algorithms while being restricted to very simple arithmetic operations in order to maintain the encryption of the data. In addition, each operation on encrypted data is very expensive, so it is necessary to minimize the number of operations in order to have a feasible model.

Achievements (according to KPIs)

  1. Implementation of the new algorithm.
  1. Testing the new algorithm on different datasets and comparing it to the classical algorithm.
  2. Writing up part of the results section in a paper.

Share this post

Share on facebook
Share on twitter
Share on linkedin
Share on email