Fully funded PhD postion in Deep Generative Models and Missing Data at DTU in Copenhagen
Jes Frellsen
Hi all,
I currently have a vacant PhD position in my group on Deep Generative Models and Missing Data. The position is fully funded, including tuition fee, travel and salary/pension.
What?
Building on our previous work in this area [1,2], the objective of this project is to study and develop methods for learning deep generative models that jointly learn the data generating process and missing-data mechanism in MNAR data. We will consider models based on Variational Autoencoders (VAEs), normalizing flows and Generative Adversarial Networks (GANs). Of particular interest is advanced inference techniques, heterogeneous data and how to make minimal but principal modelling assumptions about the missing process, but the focus will depend on the chosen candidate. We are interested in various applications within image analysis, collaborative filtering and bioinformatics.
Where?
You will be working in the Section for Cognitive Systems at the Technical University of Denmark, which is an internationally renowned group for machine learning research. Both salary and working conditions are excellent. We emphasize a healthy work/life balance based on the premise that you do the best work when you are happy. The group is a down-to-earth and fun place to be. Most group members live in Copenhagen, which is often named as one of the best city in the world to live.
The project includes collaborations and stays with Professor Julie Josse at École Polytechnique and Research Scientist Pierre-Alexandre Mattei at INRIA in Sophia Antipolis.
How?
More information regarding this position and how to apply can be found at:
https://www.compute.dtu.dk/om_os/ledige_stillinger/job?id=4aadce33-5a31-4f81-a43e-496b55a7b9fa
The application deadline is 12 August 2020.
Sincerely,
Jes Frellsen
[1] Mattei P-A, Frellsen J (2019) MIWAE: Deep Generative Modelling and Imputation of Incomplete Data. Proceedings of the 36th International Conference on Machine Learning (ICML 2019).
[2] Ipsen NB, Mattei P-A, Frellsen J (2020) not-MIWAE: Deep Generative Modelling with Missing not at Random Data. arXiv: 2006.12871