I have a PhD position to design theoretically grounded
online/reinforcement learning algorithms for carbon and cost aware
scheduling of computational jobs to reduce CO2 emissions. The aim
is to find ways to schedule computation at times of low carbon
intensity of electricity supply, without relying on predictability
of carbon intensity. (Due to presence of multiple independent
consumers on the network, prediction of carbon intensity is
impossible.)
Candidates are expected to have solid theoretical background and
knowledge of online and reinforcement learning.
Supervisors: Yevgeny Seldin and Raghavendra Selvan.
Apply at
https://di.ku.dk/english/about/vacancies/three-phd-fellowships-in-machine-learning/
When applying write down "Yevgeny Seldin" in the "Principal
Supervisor" field.
(The project is not listed explicitly in the call, because the
position was opened after the call for the other three PhD
positions mentioned in the call was published, but all
applications will be considered.)
Application deadline: 15 January 2025
Contact details for questions: Yevgeny Seldin
<sel...@di.ku.dk>