The Division of Statistics and Machine Learning (STIMA) at Linköping University is recruiting several Postdocs in Machine Learning, to work on a range of research topics as described below. Linköping University is one of the leading AI institutions in Sweden and is well recognized by solid contributions to top machine learning conferences. The division has strong links to prominent national research initiatives, such as WASP and ELLIIT.
The goal of this recruitment is to expand our research activities related to several topics in machine learning, both applied and more fundamental. Common to all research topics is that you will work in an academic environment characterized by a tight integration of statistics and machine learning. We have a wide network of strong international collaborators all around the world, for example at the University of Cambridge, University of Oxford, Columbia University, UC Berkeley, University of Amsterdam, and University of British Columbia, giving rise to excellent opportunities for international collaboration. Furthermore, you will have access to state-of-the-art computing infrastructure for machine learning (e.g. through BerzeLiUs). The contact person at the division (PI/co-PI) for all projects listed below is associate professor Fredrik Lindsten.
Click here to apply. We anticipate several recruitment rounds during 2022. The current application deadline is: February 14
The suggested research topics for the open positions are:
1. Machine learning for materials discovery
You will be part of the KAW-funded cross-disciplinary project The 2D Materials Frontier. In collaboration with physicists in Linköping and Uppsala, we will develop novel machine learning methods to aid in the quest for new ultra-thin materials, with applications in energy storage, catalysis, and water treatment, to mention a few. Relevant machine learning competences include graph neural networks, geometric deep learning, and active learning.
2. Novel AI methods for experimentally constrained protein structure prediction
This position is part of a joint collaboration between the two largest research programs in Sweden, the Wallenberg AI, Autonomous Systems and Software Program (WASP) and the SciLifeLab and Wallenberg National Program for Data-Driven Life Science (DDLS), with the ultimate goal of solving ground-breaking research questions across disciplines.
In collaboration with Prof Sebastian Westenhoff at Uppsala University, we will develop novel algorithms to include instance-specific experimental constraints in machine learning models, effectively bridging the gap between AI predictions and experimental observation. The algorithms will be widely applicable in many areas of AI. However, in particular we will focus on combining machine-learning-based protein structure predictions with experimental constraints obtained by single-particle cryo EM, to improve structure prediction and characterization of conformational heterogeneity of proteins. Relevant machine learning skills include graph neural networks, geometric deep learning, transformers, energy-based models, ensembles, and Monte Carlo methods.
The LiU postdoc will represent the Wallenberg AI, Autonomous Systems and Software Program (WASP) in this cross-disciplinary project. WASP is Sweden’s largest individual research program ever, a major national initiative for strategically motivated basic research, education and faculty recruitment. The program addresses research on artificial intelligence and autonomous systems acting in collaboration with humans, adapting to their environment through sensors, information and knowledge, and forming intelligent systems-of-systems. The vision of WASP is excellent research and competence in artificial intelligence, autonomous systems and software for the benefit of Swedish industry.
The project is done in close collaboration with Westenhoff’s lab and a DDLS-funded postdoc in Uppsala. The SciLifeLab and Wallenberg National Program for Data-Driven Life Science (DDLS) is a 12-year initiative that focuses on data-driven research, within fields essential for improving the people’s lives, detecting and treating diseases, protecting biodiversity and creating sustainability. The programme will train the next generation of life scientists and create a strong computational and data science base. The program aims to strengthen national collaborations between universities, bridge the research communities of life sciences and data sciences, and create partnerships with industry, healthcare and other national and international actors.
3. AI-powered carbon border adjustments
The project is part of the Vinnova initiative AI in the service of climate. In a collaboration between climate policy experts and machine learning researchers at Linköping University, 2050 Consulting, Toyota Material Handling and Alfa Laval, we will develop a machine learning system for automatically generating and validating life cycle assessments. The goal of the project is to enable trustworthy carbon border adjustments, with the potential of significantly reducing CO2 emissions. Relevant machine learning skills include Bayesian inference, probabilistic graphical models, text analysis and natural language processing. Read more here.
4. Uncertainty quantification for machine learning
In this more basic research project, we will develop novel theory and methods for quantifying and handling various types of uncertainty in machine learning. This includes, for instance, interpretable calibration evaluation for generic probabilistic models, epistemic uncertainty quantification for neural networks, robust methods for out-out-distribution detection, and probabilistic modeling of input dependencies.
5. Spatio-temporal data analysis and hybrid modeling
Processes that evolve over both time and space are ubiquitous in science. Examples include the spread of a disease in a region, or the flow of traffic and evolution of various variables pertaining to the risk of traffic accidents. We will work on developing a new class of hybrid models for spatio-temporal data analysis and prediction, where complex simulators are integrated with machine learning at an unprecedented level.
There is a joint opening for all topics listed above. Please specify in your application letter which project(s) that you would be interested in working on - it is possible to apply for multiple topics with a single application. Alternatively, if you have other research ideas that you would like to pursue as a postdoc at STIMA, feel free to list them in your application as well – we are open to own initiatives!
Teachers’ exemption (Lärarundantaget)
Researchers and academic staff at Swedish universities retain the rights to their own research results, even when these are patentable. This system is known as the “teachers’ exemption” (Lärarundantaget) and differs from the system at universities in other countries. LiU provides support for researchers as inventors, with a unit at the university that helps researchers to patent ideas, and may even fund the patent application, while the patent is still owned by the researcher.