Dr Tom Rainforth

Senior Researcher in Statistical Machine Learning

Biographical sketch

I am currently a Senior Researcher in Statistical Machine Learning and Leader of the RainML (rainml.uk) Research Lab.  Prior to this, I was a Junior Research Fellow in Computer Science at Christ Church College, and before that I was a postdoc working with Yee Whye Teh.  I originally studied Mechanical Engineering (MEng) at the University of Cambridge, while I did my DPhil in Oxford under the supervision of Frank Wood and Michael Osborne, focusing mostly on probabilistic programming and Monte Carlo methods.  I also had a stint working in the Ferrari Formula 1 team in between the two periods of study.

Personal website: https://www.robots.ox.ac.uk/~twgr/

Research Interests

My research covers a wide range of topics in and around statistical machine learning and experimental design, with areas of particular interest including: 

  • Bayesian experimental design
  • Probabilistic and data-efficient approaches to machine learning
  • Active learning
  • Deep learning, with a particular focus on probabilistic approaches, deep representation learning, and deep generative models
  • Probabilistic programming
  • Approximate inference and Monte Carlo methods

Please see my Google Scholar page for an up-to-date list of publications.

Publications

Barrett, B., Camuto, A., Willetts, M. and Rainforth, T. (2022) “Certifiably Robust Variational Autoencoders”, in Proceedings of Machine Learning Research, pp. 3663–3683.
Joy, T., Schmon, S., Torr, P., Narayanaswamy, S. and Rainforth, T. (2021) “Capturing label characteristics in VAEs”, in Proceedings of the International Conference on Learning Representations (ICLR 2020). OpenReview.
Camuto, A., Willetts, M., Roberts, S., Holmes, C. and Rainforth, T. (2021) “Towards a theoretical understanding of the robustness of variational autoencoders”, in. Journal of Machine Learning Research, pp. 3565–3573.
Ivanova, D., Foster, A., Kleinegesse, S., Gutmann, M. and Rainforth, T. (2021) “Implicit Deep Adaptive Design: Policy-Based Experimental Design without Likelihoods”, in Advances in Neural Information Processing Systems, pp. 25785–25798.
Kossen, J., Band, N., Lyle, C., Gomez, A., Rainforth, T. and Gal, Y. (2021) “Self-Attention Between Datapoints: Going Beyond Individual Input-Output Pairs in Deep Learning”, in Advances in Neural Information Processing Systems, pp. 28742–28756.
Tolpin, D., Zhou, Y., Rainforth, T. and Yang, H. (2021) “Probabilistic Programs with Stochastic Conditioning”, in Proceedings of Machine Learning Research, pp. 10312–10323.
Foster, A., Ivanova, D., Malik, I. and Rainforth, T. (2021) “Deep Adaptive Design: Amortizing Sequential Bayesian Experimental Design”, in Proceedings of Machine Learning Research, pp. 3384–3395.
Kossen, J., Farquhar, S., Gal, Y. and Rainforth, T. (2021) “Active Testing: Sample-Efficient Model Evaluation”, in Proceedings of Machine Learning Research, pp. 5753–5763.
Farquhar, S., Gal, Y. and Rainforth, T. (2021) “ON STATISTICAL BIAS IN ACTIVE LEARNING: HOW AND WHEN TO FIX IT”, in ICLR 2021 - 9th International Conference on Learning Representations.
Foster, A., Pukdee, R. and Rainforth, T. (2021) “IMPROVING TRANSFORMATION INVARIANCE IN CONTRASTIVE REPRESENTATION LEARNING”, in ICLR 2021 - 9th International Conference on Learning Representations.

Contact Details

Email: rainforth@stats.ox.ac.uk

Office: 1.06

Pronouns: He/Him

Research Groups

Computational Statistics and Machine Learning

Graduate Students

Alex Forster
Angus Phillips
Freddie Bickford Smith
Guneet Singh Dhillon
Andrew Campbell
Desi Ivanova
Jannik Kossen
Ning Miao
Tim Reichelt
Mrinank Sharma
Yuyang Shi
Shahine Bouabid
Jin Xu