- Associate Professor in Machine Learning
I'm interested in the principles of machine learning, with a particular focus on modern deep learning methods. My research falls into the following themes:
- Unsupervised Representation Learning: What are the hallmarks of good neural representations of data, and how do we discover these from data without labels? How can we formalize the goals and principles of unsupervised representation learning, of self-supervised learning, of transfer learning? Can we understand why self-supervised learning works so well in vision and NLP?
- Optimization and Generalization in Deep Learning: Why do deep networks generalise? What is the role of stochastic gradient descent? What is the role of the (over)parametrization of neural networks? Are natural gradient descent methods a good idea for neural network training, from the perspective of generalization?
- Probabilistic foundations: How should we represent uncertainty about models in deep learning? Is Bayesian inference a good principle for deep learning? Can we develop alternative ways of quantifying uncertainty that line up better with our goal? Can we develop practical inference algorithms in prediction-oriented, loss-calibrated situations?
- Causally Robust Deep Learning: How can we devise principled methods that learn causally meaningful representations from observed data, which robustly generarlise under distribution shift or changing policies? Can we identify invariant causal mechanisms using deep learning?
Biography
I finished my PhD in Machine Learning at the Cambridge University Engineering Department in 2013. Back in those days I worked on Bayesian inference, nonparametric and kernel methods. Since then, I worked in the London technology sector: in various tech startups and briefly in venture capital. I served as Principal Research Scientist at deep learning startup Magic Pony Technology, where we focussed on applying machine learning to the problem of lossy image and video compression. Following the acquisition of Magic Pony by Twitter in 2016, I served as Senior Machine Learning Researcher at Twitter where I worked on a range of ML-related projects including computer vision, recommender systems and ML ethics and fairness. I joined the Department of Computer Science and Technology in 2020 as a Senior Lecturer. I continue to advise Twitter's ML Ethics, Transparency and Accountability (META) team as a staff research scientist.
Teaching
- Theory of Deep Learning (R252)
- graduate module for Part III CST (4th year undergraduate) and MPhil Advanced Computer Science
- Deep Learning and Neural Networks (DeepNN)
- Part II CST (3rd year undergraduate)
- Advanced Topics in Machine Learning or Natural Language Processing (R250)
- graduate module for Part III CST (4th year undergraduate) and MPhil Advanced Computer Science
- Topic 6: Causal Inference
Publications
For a list of publications, please refer to my google scholar page.