I work on topics in machine learning theory and related areas, including game theory, information theory, statistics and optimisation. My current interests include pure exploration in multi-armed bandit models, game tree search, and provably accelerated learning in statistical and individual-sequence settings, aka learning faster from easy data. I am also interested in hypothesis testing using martingales.
- Online learning, regret minimization, online convex optimization, bandits
- Pure exploration
- Connections between learning and games
- Sequential hypothesis testing, theory of e-values and martingales.