Distributed and Privacy-Preserving Computations in R

Balasubramanian Narasimhan

Department of Biomedical Data Sciences and Department of Statistics

Stanford University

January 10-12, 2018

IPAM, UCLA

Abstract

The R programming environment provides a rich suite of methods and algorithms for statistical model fitting and computation in general. The exploratory interface of R makes it an ideal platform for prototyping, analyzing and deploying new algorithms in an open and extensible manner. This tutorial will begin with a short self-contained introduction to R and its package mechanism for programming extensions. We will then discuss some packages for distributed, cryptographic, homomorphic, and differential privacy computations. The tutorial will include hands-on programming exercises. The goal is to provide the background and programming knowledge so that participants can develop their own packages implementing new algorithms.

Requirement

A reasonably modern laptop with internet connectivity.