Department of Computer Science and Technology

Fixed-term: The funds for this post are available for 2 years.

We are looking for a Research Associate to work on the project Machine Learning Approaches as Aerospace Engineering Assistants. Under the supervision of Dr. Carl Henrik Ek and Dr. Henry Moss the postholder will be responsible to develop of probabilistic models and approximate inference methods. Specifically the postholder will work in a collaborative project between the Department of Computer Science and Boeing. The project is focused on building probabilistic models for the task of engineering designs. The design process is very labour intensive and is fundamentally driven by tacit knowledge that designers build up after many years of practice. This knowledge is reflected in the data of the resulting designs. In this project we aim to exploit the statistics in this data to build tools that allow us to shed light on design process. In specific we aim to build probabilistic models that allows for constrained and conditional sampling of designs and allows for optimisation and exploration of designs. The postholder will also be expected to be involved in supervision of PhD and MPhil students working on projects that are part of the same collaboration. Candidates will have completed a PhD within the field of Machine Learning with a specific focus on probabilistic and Bayesian models. In specific we are looking for candidates with expertise on Bayesian optimisation, experimental design, Gaussian processes and active sampling. Given that the engineering design process consists of many steps the ideal candidate have experience in hierarchical and composite probabilistic models.

https://www.jobs.cam.ac.uk/job/44781/