Researcher on Reinforcement Learning

THIS IS A RE-ADVERTISEMENT: PREVIOUS CANDIDATES NEED NOT APPLY



A full-time postdoctoral position is available for a researcher in the area of
reinforcement learning. This is a European Research Council-funded project,
working with Professor Shimon Whiteson. The goal of the project is to develop
a new class of reinforcement learning and sample-based decision-theoretic
planning methods that overcome fundamental obstacles to the efficient
optimisation of control policies for autonomous agents. Creating agents that
are effective in diverse settings is a key goal of artificial intelligence
with large potential implications in robotics, e-commerce, information
retrieval, traffic control, etc.



The postholder will be expected to collaborate in the preparation of research
papers, development and analysis of new algorithms, present papers at
conferences, act as a source of information and advice to other members of the
group on scientific procedures and experimental techniques, which may include
stochastic optimisation, Bayesian optimisation, Bayesian quadrature, and deep
learning.



The postholder will require a doctoral degree (or be very close to completion)
in computer science or a related area, together with a documented track record
of published, peer-reviewed research in machine learning and/or decision-
theoretic planning (or a related area) and possess strong mathematical skills
in probability and statistics together with strong programming skills.
Experience of conducting large-scale experiments with complex datasets and
simulators is highly desirable.



The closing date for applications is 12.00 noon on 1 December 2016.