Brian McWilliams

About me.

I am a research scientist at Disney Research Zurich heading the Analytics group. I am particularly interested in deep learning, randomized algorithms and optimization for large scale learning.

I completed my Ph.D. in April 2012 in the Statistics Section of the Department of Mathematics Imperial College, London under the supervision of Giovanni Montana . Previously, I did the MSc. in Informatics specialising in Machine Learning, Neuroinformatics and Intelligent Robotics at the University of Edinburgh and a BEng. in Computer Systems Engineering at the University of Warwick.

Until August 2015 I was a postdoctoral researcher and lecturer in the Institute of Machine Learning at ETH Zurich.

Contact details.

Open positions: We have postdoc and internship openings for talented researchers in the fields of machine learning, optimization and statistics. Particularly in the areas of deep learning for natural language understanding and reinforcement learning. If that sounds like your thing, please get in touch.


August 2016

RadaGrad! our paper on scalable approximations to full-matrix AdaGrad using random projections is accepted to NIPS 2016!

December 2015

Dual-LOCO is accepted to AISTATS 2016 as an oral presentation!

June 2015

  • In September 2015 I will be heading down the road (it is a small world, afterall) to join Disney Research Zurich as a research scientist in the analytics group.

  • Christina Heinze has developed a great SPARK package for Loco and Dual-Loco. More information is available at the following Github page.
  • Papers.


  • Preserving Differential Privacy Between Features in Distributed Estimation .
    Christina Heinze, Brian McWilliams, Nicolai Meinshausen.
  • Semi-Supervised Learning with Correlated Convolutional Networks .
    Ivan Ovinnikov, Brian McWilliams, Dmitry Laptev, Joachim Buhmann.
  • LOCO: Distributing Ridge Regression with Random Projections.
    Christina Heinze, Brian McWilliams, Nicolai Meinshausen and Gabriel Krummenacher. arXiv.
    Software (maintained by Christina).
  • A Variance Reduced Stochastic Newton Method.
    Aurelien Lucchi, Brian McWilliams and Thomas Hofmann. arXiv.

  • 2016

  • Scalable Adaptive Stochastic Optimization Using Random Projections.
    Gabriel Krummenacher, Brian McWilliams, Yannic Kilcher, Joachim Buhmann and Nicolai Meinshausen. NIPS 2016.
  • A Benchmark Dataset and Evaluation Methodology for Video Object Segmentation.
    Federico Perazzi, Jordi Pont-Tuset, Brian McWilliams, Markus Gross, Luc Van Gool and Alexander Sorkine-Hornung. CVPR 2016. Project page.
  • DUAL-LOCO: Distributing Statistical Estimation Using Random Projections.
    Christina Heinze, Brian McWilliams and Nicolai Meinshausen. AISTATS 2016. arXiv. Software (maintained by Christina).

  • 2015

  • Variance Reduced Stochastic Gradient Descent with Neighbors.
    Thomas Hofmann, Aurelien Lucchi, Simon Lacoste-Julien and Brian McWilliams. NIPS 28. arXiv.
  • DUAL-LOCO: Preserving privacy between features in distributed estimation.
    Christina Heinze, Brian McWilliams and Nicolai Meinshausen. NIPS WS on Learning and privacy with incomplete data and weak supervision.
  • Learning Representations for Outlier Detection on a Budget.
    Barbora Micenková, Brian McWilliams and Ira Assent. arXiv.

  • 2014

  • RadaGrad: Random Projections for Adaptive Stochastic Optimization.
    Gabriel Krummenacher and Brian McWilliams. 7th NIPS Workshop on Optimization for Machine Learning. pdf.
  • Fast and Robust Least Squares Estimation in Corrupted Linear Models.
    With Gabriel Krummenacher, Mario Lučić and Joachim M. Buhmann. NIPS 27. arXiv. Software (maintained by Gabriel). Slides and video from the Zuri ML meetup.
  • Learning Outlier Ensembles: The Best of Both Worlds – Supervised and Unsupervised.
    Barbora Micenková, Brian McWilliams and Ira Assent. KDD Workshop on Outlier Detection & Description under Data Diversity (ODD²). pdf.
  • Subspace clustering of high-dimensional data: a predictive approach.
    With Giovanni Montana. Data Mining and Knowledge Discovery. 28(3): 736-772. arXiv.

  • 2013

  • Correlated random features for fast semi-supervised learning.
    With David Balduzzi and Joachim M. Buhmann. In Advances in Neural Information Processing Systems (NIPS) 26. arXiv. Matlab code. Poster.
  • Pruning random features with correlated kitchen sinks (1 page abstract).
    With David Balduzzi. SPARS 2013.

  • 2012 and earlier


    Courses I am involved with.

    Fall Semester 2014

    I co-lectured Probabilistic Graphical Models for Image Analysis with Dr. Aurelien Lucchi.

    Spring Semester 2014

    I was head TA of Computational Intelligence Laboratory. This course is now taught by Prof. Hofmann and headed by Martin Jaggi and Aurelien Lucchi.


    The website for the probabilistic graphical models course that Dr. David Balduzzi and I taught at Uni Basel in summer 2013 is located here.



    Disney Research Zurich
    Stampfenbachstrasse 48
    8006 Zurich