Research Associate in Machine Learning

Prior work in our group (Hartley & Tsaftaris, arXiv 2022 and Jegorova et al, arXiv, 2021) has shown that supervised AI models are very susceptible to spurious correlation even when it is extremely rare. We don’t know how the representations are affected, or how shortcut learning contributes, or whether this is a pure artifact of stochastic training. 

We are looking for an enthusiastic postdoctoral researcher to join our team led by Prof. Sotirios Tsaftaris in a prestigious high-profile project (funded by EPSRC) to help us address some of these questions and in the meantime develop new generative representation learning algorithms. We are keen to also investigate causal representation learning as a mechanism for discerning information that originates from imperfect data collection processes (which can create the spurious correlation).  

Areas of particular interest thus include generative models, representation learning, causal representations, concept learning, learning theory. 
The candidate will:

• join a dynamic, international team of several postdocs and students working in representation learning and causality at the University of Edinburgh –a powerhouse in artificial intelligence, engineering and medicine;
• contribute to an exciting project where deep learning meets rare data directly addressing industry and societal needs;
• visit and interact with industry collaborators in healthcare (Canon Medical Research Europe), in AI (Amazon Web Services) and ELLIS Units across Europe; and
• co-organise workshops on representation learning.

What we are looking for: A relevant PhD; Expertise in the areas of interest above evidenced by related publications in well-regarded venues; An ability to deliver quick proof-of-concepts of your original thinking; Good communication skills and ability to work within a team; and 
a positive attitude illustrating enthusiasm in high risk/reward open-ended research.

This post is available immediately with a start no later than April 1st, 2023. Informal enquiries to Prof. Tsaftaris (s.tsaftaris@ed.ac.uk; https://vios.science) are welcome quoting this project and with a CV. 
Deadline to apply: 21/Oct/2022, 5pm UK

Click here to apply and for more information:
https://elxw.fa.em3.oraclecloud.com/hcmUI/CandidateExperience/en/sites/CX_1001/job/5238 

 
Edinburgh has been consistently voted a top place to live and work and most recently ranked #1 in the world by Time Out.