Les commentaires sont clos.

Ph.D. position in information science - Taming Inverse Problems over the Continuum

12 Février 2023

Catégorie : Doctorant

Open Ph.D. position on the theoretical foundations of information science
Taming Physical Inverse Problems over the Continuum


  • Theoretical data science
  • Estimation and statistical learning
  • Signal processing
  • Numerical methods

Research scope

Inferring on statistical properties of probability distributions from their empirical samples or recovering highly structured signals from their coarse and distorted measurements are ubiquitous, yet challenging tasks in signal processing and machine learning. Those problems find myriad applications in the area ofdata and experimental sciences, including data classification, distribution learning, optical and radar imaging, astronomy, telecommunication, and the identification of neural recordings. From a mathematical perspective, those inference problems can often be translated in a versatile fashion as to reconstruct a continuously valued measure from low-dimensional observations.

The general research scope of the Ph.D. project is at the intersection of Data Science, Information Theory, and Theoretical Machine Learning. It will aim to investigate the statistical feasibility of specific continuously valued inverse problems and to develop numerical methods to tackle modern problemsarising from physics, imaging and communication, and network modalities. A particular attention will be placed on algorithmic scalability to large data volume and robustness to low-quality inputs. Applicative areas include but are not limited to:

-Learning of mixtures: A statistical problem at the core of data classification.

-Physical layer security: A recent information-theoretic framework that seeks to leverage the structural properties of wireless communication channels to encrypt the transmitted information.

-Super-resolution imaging and radar: A signal processing problem that aims at recovering the parameter of a structured signal in a highly degraded environment with optimal precision by harnessing prior information.

During the Ph.D. project, the successful applicant will be strongly encouraged to develop and strengthen his/her own research interests and to construct his/her personal research plan within this generic scope.


Possibility to start the Ph.D. with an internship on spring/summer 2023.

More information about this subject, and the application process on