We are looking for a highly creative and motivated PhD candidate to join the Data Mining Group at Eindhoven University of Technology. The candidate will be supervised by Dr. Vlado Menkovski, PI of the project and work within the Immuno-engineering research program, which is a collaboration with the department of Biomedical Engineering.
Machine Learning models, particularly deep neural networks, have shown outstanding capabilities to deliver accurate prediction in high dimensional settings [1, 2]. However, the goal of the scientific inquiry goes beyond prediction and aims at explanations that can be integrated with the existing knowledge.
In the context of data-driven scientific discovery, there are certain unique challenges. Commonly a vast amount of domain knowledge is available that can strongly benefit the development of deep learning models particularly when data is scarce and costly to obtain.
Therefore, in this project we aim to go beyond the Supervised Learning formulation of Machine Learning to Generative Models with rich structure that can capture and express the relationship between meaningful factors and the observed data. An example of such approaches would be the many flavors of conditional Variational Autoencoders that can be used for both simulation (data generation) and parameter estimation (estimate the values of factors of interest given the observations) [3, 4]. We will also study different approaches to express the known constraints and symmetries in the data generation process, particularly drawing inspiration from developments in directions such as Physic- informed neural networks.
Being part of the Immuno-engineering program, the work will be in close collaboration with researchers in Biomedical Engineering and related fields. With-in the scope of the program, problems and goals are developed and data is collected.
This PhD position provides ideal conditions for developing yourself as a researcher in the emerging field of Scientific Machine Learning. The position is embedded in a strong and growing Machine Learning community in the Mathematics and Computer Science Department. It is part of a project working on real-world and highly impactful problems in Immuno-engineering that also provides access to leading researchers in that field. As such we aim to both publish and advance the state of the art in leading Machine Learning as well as Immuno-engineering and Biomedical engineering venues.
1. Corbetta, A., Menkovski, V., Benzi, R., Toschi, F. "Deep learning velocity signals allows to quantify turbulence intensity." Science Advances, eaba7281 (2021).
2. Matos, F., V. Menkovski, F. Felici, A. Pau, F. Jenko, TCV Team, and EUROfusion MST1 Team. "Classification of tokamak plasma confinement states with convolutional recurrent neural networks." Nuclear Fusion 60, no. 3 (2020): 036022.
3. Perez Rey, L. A., Menkovski, V., & Portegies, J. W. Diffusion Variational Autoencoders. IJCAI., Yokohama, Japan (2020).
4. Wang, Liwei, Yu-Chin Chan, Faez Ahmed, Zhao Liu, Ping Zhu, and Wei Chen. "Deep generative modeling for mechanistic-based learning and design of metamaterial systems." Computer Methods in Applied Mechanics and Engineering 372 (2020): 113377. 113377
We are looking for a motivated candidate with:
Do you recognize yourself in this profile? For any further inquiries on the content of the position,
please contact Vlado Menkovski (v.menkovski[at]tue.nl) prefixing the email title with “[Immuno-engineering Vacancy]”.
For information about terms of employment, click here or contact HRServices.MCS[at]tue.nl.
Please visit www.tue.nl/jobs to find out more about working at TU/e!
We invite you to submit a complete application by using the 'apply now'-button on this page.
The application should include a:
We look forward to your application and will screen it as soon as we have received it.
Screening will continue until the position has been filled.
We do not respond to applications that are sent to us in a different way.
Please keep in mind you can upload only 5 documents up to 2 MB each. If necessary please combine files.