We are looking for a PhD student to study the mechanisms of information transfer and feature learning in biological neural networks.
IDLab is a research group of UGent, as well as a core research group of imec. IDLab mainly performs fundamental and applied research on data science and internet technology, and counts over 300 researchers.
The AIRO team of IDlab, active since the beginning of this century, has its roots in artificial neural networks and robotics with a tendency for blue sky approaches. As one of the original groups involved in reservoir computing, AIRO has a historical perspective of what can be achieved by exploiting the natural dynamics of physical systems. This vision has since been applied to “physical reservoir computing”, e.g., by computing with analog (active or passive) photonic devices or with compliant robotic bodies. However, although it holds a lot of promise w.r.t. power- and resource-efficient computing, the concept of reservoir computing by itself is not powerful enough to compete with state-of-the-art non-algorithmic computation for complex tasks. One of the main reasons for this is that there is no systematic design methodology (as in digital computing), nor a solid learning-based approach (as in artificial neural networks). The field of AI is currently dominated by various brands of deep learning. That, in turn, largely relies on supervised learning, which is notoriously data-hungry and compute-hungry. despite the huge amounts of data and compute-power that are currently being invested, many tasks that are easy for humans remain incredibly hard for even the most powerful AI systems. For many researchers and companies, the answer to this remains “more data” and “more GPUs” (especially for those selling data and GPU-time as a business model). In our team, we are convinced that the answer lies in deepening our understanding of why our brains seem to be more efficient at learning such complex tasks. Many such research directions are currently being explored, for example, in concepts such as “grounding” or “curriculum learning”. In this process, the focus lies on finding more efficient and biologically inspired ways of learning features and feature hierarchies.
Through our involvement in the SmartNets Marie Curie European Training Network, we are looking for a PhD student to study the mechanisms of information transfer and feature learning in biological neural networks. In order to establish more efficient learning in artificial neural networks, we will try to identify mechanisms that are both useful and transferable to digital computing hardware.
You will be enrolled at Ghent University for a PhD in Computer Science Engineering. However, your research will be highly interdisciplinary. You will need to combine in-depth understanding of biological learning, artificial learning and its efficiency as a hardware implementation. As PhD student at Ghent university, you will collaborate with enthusiastic colleagues at IDLab-AIRO and our international partners in the SmartNets project. As an ESR in the SmartNets network, you will form an active training network with the other ESRs in the project and you are required to spend part of your PhD time (~ 2 times 3 months) with some of our partners.
- ETN offers funding for early-stage researchers only. To be eligible for recruitment within an ITN project, you therefore must – at the date of recruitment – be within the first four years (full-time equivalent research experience) of your research career and not have a doctoral degree (this means that you have obtained the first diploma which gives you access to a PhD in Computer Science engineering less than 4 years ago).
- ETN is a researcher mobility programme. You are therefore required to undertake transnational mobility in order to be eligible for recruitment in an ITN project. As such, you must not have resided in or carried out your main activity (e.g. work, studies) in the country where you have been recruited “for more 4 than 12 months in the 3 years immediately before your recruitment date”
Degree and background
- You have the degree of Master of Science, preferably in Computer Science (engineering), Cognitive or computational Neuroscience or related
- Ideally, you have a background in both, computer science (in particular, artificial neural networks and/or Bayesian networks), (computational) neuroscience and (properties of) physical realisations (e.g., trade-offs and efficiency in digital hardware)
- Your degree must be equivalent to 4 or 5 years of studies (bachelor + master) in the European Union, you must have a solid academic track record (graduation cum laude or grades in the top 15% percentile)
- You are fluent in written and spoken English
- You speak and understand Dutch or are willing to learn Dutch during your time at our lab
- You are creative and prefer to find “out-of-the-box” solutions
- You are particularly interested in blue-sky fundamental research, while keeping practical applicability in mind
- You are interested in and motivated by the research topic, as well as in obtaining a PhD degree.
- You are systematic and organized, have excellent analytical skills, and can work independently as well as in team.
We offer a fully funded PhD scholarship for a period of 3 years (upon positive progress evaluation), possibly extendable with at most 1 year. The PhD research mainly has fundamental and innovative aspects. You will join a young and enthusiastic team of researchers, post-docs and professors. This PhD position is available immediately. More information on ESR-positions can be found here
Apply with extensive motivation letter, focusing on matching your background and insights with the content of this project, scientific resume, diplomas and detailed academic results (courses and grades), English proficiency scores for candidates with degrees outside the EU, relevant publications (if any), and two reference contacts. Incomplete applications will not be considered! Send your application to prof. dr. ir. Joni Dambre (Joni.Dambre@UGent.be). After the first screening, suitable candidates will be invited for an interview (also possible via Skype) and may get a skills assignment.
Closing date and timeline
This position is open immediately and the contract has to start ideally, on or before March 1st, 2021. Candidates who are not yet available (e.g., because they haven’t graduated yet) will not be considered! For selections, we will use a 2 step procedure:
- We will begin our selection procedure on January 4th 2021. Do not expect any answers to your application before that date. Promising candidates will be invited for a video call interview.
- Any applications we receive after January 4th, 2021 and before February 15th, 2021 will be considered as long as we haven’t found a suitable candidate.