niekbouman.github.io

Niek J. Bouman, PhD

Researcher in Secure Multiparty Computation and Privacy Preserving Machine Learning, funded by H2020 project Scalable Oblivous Data Analytics (SODA)

n . j . bouman (at) tue . nl

Current Affiliation

TU/e logo

Technische Universiteit Eindhoven

My research focuses on bridging multiparty computation and machine learning, in particular on

secure linear algebra over finite fields and over the rationals
secure comparison protocols
fast implementations of big-integer arithmetic, and its formal verification of correctness
“constant-time” implementations

Upcoming Preprints

Fast Secure Comparison for Medium-Sized Integers and Its Application in Binarized Neural Networks, accepted at RSA Conference 2019, Cryptographer’s Track, together with Mark Abspoel, Berry Schoenmakers and Niels de Vreede.

We expect to upload the preprint by December

Recent Preprints

Average Point Pursuit using the Greedy Algorithm: Theory and Applications together with Andrey Bernstein (National Renewable Energy Laboratory, Golden, CO, USA)

This work is related to the work I did at EPFL on the topic of control of electrical grids in the renewable energy era

New Protocols for Secure Linear Algebra: Pivoting-Free Elimination and Fast Block-Recursive Matrix Decomposition together with Niels de Vreede

Recent Organizational Activities

workshop flyer

Workshop Data Mining with Secure Computation @ ECML PKDD 2017

Past Affiliations

ABN AMRO Research in Automatic Fraud Detection
EPFL, Lausanne, Smart Electricity Grids (animation video that explains current issues with legacy electricity grids, and presents the COMMELEC project)
Centrum Wiskunde & Informatica, Amsterdam, Quantum Cryptography (old website and thesis)

Software

See my GitHub account