Organisers: Quentin Berthet (University of Cambridge and The Alan Turing Institute, UK); Varun Kanade (University of Oxford and The Alan Turing Institute, UK)
Date: 5 October 2017
Time: 10:00 – 17:00
Venue: The Alan Turing Institute
Click here to register.
This workshop will bring together experts to talk about advances in online learning, a framework in machine learning where data is made available in a sequential manner. This one-day workshop will focus on the underlying theory and the links with applications, and will feature four talks by leading international researchers.
This framework tackles problems where an agent must perform an action at any given time, based on historical data coming in a stream. It has connections to statistics, stochastic optimisation, computer science and game theory; in recent years, algorithms and models from this field have found numerous applications in online advertising, medical trials, and utility maximization.
The field has recently attracted a lot of attention with the successes of artificial intelligence in games. It is a natural setting to gain a meaningful theoretical understanding for several learning tasks, in settings that can be stochastic (performing well with random inputs) and adversarial (in a competition or with the presence of malicious users).
Sebastien Bubeck (Microsoft Research, USA)
Nicolò Cesa-Bianchi (University of Milan, Italy)
Remi Munos (Deepmind, UK)
Jacob Abernethy (Georgia Institute of Technology, USA)
Vianney Perchet (ENS Paris-Saclay & Critereo, France)